Single-Use Technology in Biopharmaceutical Manufacture 9781119477785, 9781119477778, 9781119477839, 1119477786

Citation preview

Single‐Use Technology in  Biopharmaceutical Manufacture

Single‐Use Technology in Biopharmaceutical Manufacture Second Edition

Edited by Regine Eibl and Dieter Eibl

School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology, Zurich University of Applied Sciences, Wädenswil, Switzerland

This edition first published 2019 © 2019 John Wiley & Sons, Inc. Edition History John Wiley & Sons (1e, 2011) 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 Regine Eibl and Dieter Eibl to be identified as the authors of the editorial material in this work has been asserted in accordance with law. Registered Office John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA Editorial Office 111 River Street, Hoboken, NJ 07030, USA 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: Eibl, Regine, editor. | Eibl, Dieter, editor. Title: Single-use technology in biopharmaceutical manufacture / edited by Regine Eibl, Dieter Eibl. Description: Second edition. | Hoboken, NJ : Wiley, [2019] | Includes bibliographical references and index. | Identifiers: LCCN 2019015112 (print) | LCCN 2019017028 (ebook) | ISBN 9781119477785 (Adobe PDF) | ISBN 9781119477778 (ePub) | ISBN 9781119477839 (hardback) Subjects: | MESH: Disposable Equipment | Biopharmaceutics–instrumentation | Technology, Pharmaceutical–instrumentation | Engineering Classification: LCC RM301.4 (ebook) | LCC RM301.4 (print) | NLM QV 26 | DDC 615.7–dc23 LC record available at https://lccn.loc.gov/2019015112 Cover Design: Wiley Cover Image: Courtesy of Sartorius AG, Goettingen Set in 10/12pt Warnock by SPi Global, Pondicherry, India Printed in the United States of America 10 9 8 7 6 5 4 3 2 1

v

Contents List of Contributors  xvii Preface  xxi Part I  1

Basics  1

Single‐Use Equipment in Biopharmaceutical Manufacture: A Brief Introduction  3 Dieter Eibl and Regine Eibl

1.1 ­Background  3 1.2 ­Terminology and Features  3 1.3 ­Single‐Use Systems in Production Processes for Therapeutic Proteins such as mAbs: Product Overview and Classification  5 1.4 ­Single‐Use Production Facilities  7 1.5 ­Summary and Conclusions  7 Nomenclature  9 ­ References  9 2

Types of Single‐Use Bag Systems and Integrity Testing Methods  13 Jens Rumsfeld and Regine Eibl

2.1 ­Introduction  13 2.2 ­Bags for Fluid and Powder Handling  13 2.2.1 Tank Liners  13 2.2.2 Two‐Dimensional Bags  14 2.2.2.1 Bags for Fluid Handling  14 2.2.2.2 Bags for Powder Handling  14 2.2.3 Three‐Dimensional Bags  15 2.3 ­Bag‐Handling and Container Systems  15 2.3.1 Bag‐Handling Systems  15 2.3.2 Container Systems for in‐House Applications  17 2.3.3 Container Systems for Liquid Shipping  17 2.4 ­Single‐Use Bag Systems for Freezing and Thawing  18 2.5 ­Container Closure Integrity Testing  18 2.6 ­Summary and Conclusions  22 Nomenclature  22 ­ References  22 3

Mixing Systems for Single‐Use  25 Sören Werner, Matthias Kraume, and Dieter Eibl

3.1 ­Introduction  25 3.2 ­The Mixing Process  25 3.2.1 Definition and Description  25 3.2.2 Mixing Quality  26

vi

Contents

3.2.3 Mixing Time  26 3.2.4 Residence Time Distribution  27 3.2.5 Reynolds Number  27 3.2.6 Specific Power Input  27 3.3 ­Single‐Use Bag Mixing Systems  27 3.3.1 Overview and Classification  27 3.3.2 Mixing Systems with Rotating Stirrer  28 3.3.2.1 Levitated Mixers  28 3.3.2.2 Magnetic Mixers  29 3.3.2.3 Mixers with Sealing  30 3.3.3 Mixing Systems with Tumbling Stirrer  31 3.3.4 Mixing Systems with Oscillating Devices  31 3.3.5 Hydraulically Driven Mixing Systems  32 3.4 ­Summary and Conclusions  33 Nomenclature  33 ­ References  33 4

Single‐Use Bioreactors – An Overview  37 Valentin Jossen, Regine Eibl, and Dieter Eibl

4.1 ­Introduction  37 4.2 ­SUB History  38 4.2.1 Phase 1: Early Beginnings  38 4.2.2 Period 2: Establishment of Disposable Membrane Bioreactors, Multitray Cell Culture Systems, and the First Bag Bioreactors  38 4.2.3 Period 3: Expansion of Wave‐Mixed, Stirred, Orbitally Shaken, and Further SUB Types  40 4.3 ­Comparison of the Current, Most Common SUB Types  40 4.3.1 Wave‐Mixed SUBs  40 4.3.2 Stirred SUBs  43 4.3.3 Orbitally Shaken SUBs  46 4.4 ­Decision Criteria for Selection of the Most Suitable SUB Type  47 4.5 ­Summary and Future Trends  48 Nomenclature  48 ­References  48 5

Systems for Coupling and Sampling  53 Cedric Schirmer, Sebastian Rothe, Ernest Jenness, and Dieter Eibl

5.1 ­Introduction  53 5.2 ­Components of Single‐Use Transfer Lines  53 5.2.1 Tubes  53 5.2.2 Fittings and Accessories  54 5.2.3 Connectors  55 5.2.4 Valves and Clamps  55 5.2.5 Pumps  55 5.3 ­Systems for Aseptic Coupling  57 5.3.1 Connection under Laminar Flow  57 5.3.2 Steam‐in‐place Connection  57 5.3.3 Aseptic Coupling  57 5.3.3.1 Aseptic Connectors  57 5.3.3.2 Welding 59 5.3.4 Aseptic Transfer Systems  59 5.4 ­Aseptic Disconnection  62 5.5 ­Systems for Sampling  64 5.5.1 Single‐Use Sampling Systems for Conventional Systems  64

Contents

5.5.2 Single‐Use Sampling Systems for Single‐Use Systems  65 5.6 ­Summary and Conclusion  66 Nomenclature  66 ­ References  66 6

Sensors for Disposable Bioreactor Systems  69 Tobias Steinwedel, Katharina Dahlmann, Dörte Solle, Thomas Scheper, Kenneth F. Reardon, and Frank Lammers

6.1 ­Introduction  69 6.2 ­Interfaces for Sensor Technology  70 6.3 Considerations of Extractables and Leachables from Integrated Sensors  71 6.4 Optical Chemosensors  72 6.4.1 Overview  72 6.4.2 Optical Oxygen Sensors  72 6.4.3 Optical pH Sensors  73 6.4.4 Optical Carbon Dioxide Sensors  73 6.5 Spectroscopic Sensors  73 6.5.1 Overview  73 6.5.2 UV/VIS Spectroscopy  74 6.5.3 Infrared Spectroscopy  74 6.5.4 Fluorescence Spectroscopy  75 6.5.5 Raman Spectroscopy  75 6.6 Capacitance Sensors  75 6.7 Electrochemical Sensors  76 6.7.1 Overview  76 6.7.2 Single‐Use pH Electrode  76 6.7.3 Field‐Effect Transistors  77 6.8 Biosensors  78 6.9 Conclusions and Outlook  78 Nomenclature  79 ­ References  79 7

Bioinformatics and Single‐Use  83 Barbara A. Paldus

7.1 ­Introduction  83 7.2 ­Bioinformatics and Single‐Use  84 7.3 ­Smart Sensors  86 7.4 ­Intelligent Control Systems  87 7.5 ­Continuous Processing  88 7.6 ­Conclusions  92 Nomenclature  94 ­ References  94 8

Production of Disposable Bags: A Manufacturer’s Report  95 Steven Vanhamel and Catherine Piton

8.1 ­Introduction  95 8.2 ­Materials  95 8.2.1 Most Important Polymeric Materials Used in Disposable Bags  95 8.2.1.1 Polyethylene 95 8.2.1.2 Polypropylene 96 8.2.1.3 Ethylene Vinyl Acetate  96 8.2.1.4 Polyamide or Nylon  96 8.2.1.5 Polyethylene Terephthalate  97 8.2.1.6 Ethylene Vinyl Alcohol  97 8.2.1.7 PVDC 97

vii

viii

Contents

8.2.2 Material Properties  97 8.2.2.1 PE  97 8.2.2.2 PP  98 8.2.2.3 EVA  98 8.2.2.4 PA  98 8.2.2.5 PET  98 8.2.2.6 EVOH  98 8.2.2.7 PVDC  98 8.3 ­Film Manufacturing and Molding  98 8.3.1 Introduction  98 8.3.2 Film Manufacturing  99 8.3.2.1 Blown Film Extrusion  100 8.3.2.2 Cast Film Extrusion  101 8.3.2.3 Extrusion Lamination  101 8.3.2.4 Film Extrusion for Disposable Bags Used in Biopharmaceutical Manufacturing  101 8.3.3 Molding  102 8.3.4 Quality Insurance  104 8.3.4.1 Mechanical Tests  105 8.3.4.2 Physical Testing  105 8.3.4.3 Biological Testing  105 8.3.4.4 Material‐Dependent Tests  105 8.3.4.5 Extractables and Leachables  105 8.3.4.6 Chemical Compatibility Tests  108 8.3.4.7 Functional Tests – Assembly Test  108 8.3.4.8 Functional Test – Differential Scanning Calorimetry (DSC)  109 8.3.4.9 Sterility Tests  109 8.3.4.10 Contamination Requirements  110 8.3.4.11 Expiry Date  110 8.4 ­Bag Manufacturing  110 8.4.1 Most Important Manufacturing Processes Used in the Production of Disposable Bags  110 8.4.2 Quality Insurance  112 8.4.2.1 Control of Incoming Material  112 8.4.2.2 Release of Disposable Bags  112 8.5 ­Summary and Conclusions  113 Nomenclature  115 ­ References  116 9

Single‐Use Downstream Processing for Biopharmaceuticals: Current State and Trends  117 Britta Manser, Martin Glenz, and Marc Bisschops

9.1 ­Introduction  117 9.2 ­Single‐Use DSP Today  117 9.2.1 Benefits and Constraints of Single‐Use DSP  117 9.2.2 Trends in Single‐Use DSP  117 9.2.3 Single‐Use and Continuous DSP Platforms  118 9.3 Technologies in Single-Use DSP  120 9.3.1 Clarification  120 9.3.2 Capture and Polishing  120 9.3.3 Virus Removal  121 9.3.4 Formulation  121 9.4 ­Single‐Use Continuous Downstream Processing  121 9.4.1 Clarification  121 9.4.2 Capture and Polishing  122 9.4.3 Virus Removal  122 9.4.3.1 Continuous In‐Process Mixing and Hold  123

Contents

9.4.3.2 Plug‐Flow Reactors  123 9.4.3.3 Filtration  123 9.4.3.4 Chromatography  123 9.4.4 Formulation  124 9.5 ­Integrated and Continuous DSP  124 9.6 ­Summary and Conclusions  124 Nomenclature  124 ­ References  125 10

Application of Microporous Filtration in Single‐Use Systems  127 Christian Julien and Chuck Capron

10.1 ­Introduction  127 10.2 ­Microporous Filters  128 10.2.1 Nominal Versus Absolute Removal Ratings  128 10.2.2 Particle Retention Mechanisms  128 10.2.3 Filter Media  128 10.2.4 Membrane Filters  131 10.2.5 Depth Filters  132 10.2.6 Sterilizing‐Grade Filters  133 10.2.7 Mycoplasma Retentive Filters  134 10.2.8 Virus Retentive Filters  134 10.3 ­Filter Selection  134 10.3.1 The Need for Filter Testing  134 10.3.2 Flow Decay Studies  134 10.3.3 Meeting Process Objectives  135 10.3.4 Applications Orientation  136 10.4 ­Final Sterile Filtration  136 10.4.1 Regulatory Highlights  136 10.4.2 Serial and Redundant Filtration  136 10.5 ­Filter Integrity Testing  138 10.5.1 Regulatory Highlights  138 10.5.2 PUPSIT  138 10.5.3 Filter Integrity Tests  138 10.6 ­Filter Qualification and Validation  139 10.6.1 Regulatory Highlights  139 10.6.2 Product‐Based Tests  139 10.7 ­Summary and Conclusions  140 Nomenclature  140 ­References  140 11

Extractables/Leachables from Single‐Use Equipment: Considerations from a (Bio) Pharmaceutical Manufacturer  143 Alicja Sobańtka and Christian Weiner

11.1 ­Introduction  143 11.2 ­Regulatory Environment  144 11.2.1 Pharmacopeia Chapters  144 11.2.2 Biological Reactivity and Chemical Safety  145 11.2.3 “Pharma Grade,” “Medical Grade”  145 11.2.4 Code of Federal Regulations – Food Grade  145 11.2.5 REACH  146 11.2.6 Regulatory Responsibility Chart  146 11.3 ­The (Bio)Pharmaceutical Manufacturer’s Approach  146 11.3.1 Risk Mitigation  146 11.3.1.1 Chemical Compatibility  146

ix

x

Contents

11.3.1.2 Clearance Steps  147 11.3.1.3 Pre‐Flush 147 11.3.2 Chemical Safety Assessment  148 11.3.2.1 Extractables Profiling  148 11.3.2.2 Controlled Extractables Study  149 11.3.2.3 Sum Parameters  149 11.3.2.4 Unknown Compounds  151 11.3.2.5 Simulated‐Use Extractables Study  151 11.3.2.6 Leachables Study  152 11.3.2.7 Exposure Scenario  153 11.3.2.8 Toxicological Risk Assessment  153 11.4 ­The (Bio)Pharmaceutical Manufacturer’s Challenges  153 11.4.1 Supply Chain of Single‐Use Equipment  153 11.4.2 Cost Factor  155 11.4.3 Time Factor  155 11.4.4 Production Outsourcing and Contract Manufacturers  155 11.4.5 Life‐Cycle Management  155 11.5 ­Summary  155 11.6 ­Discussion and Outlook  156 ­ Acknowledgments  156 Nomenclature  157 ­References  157 12

The Single‐Use Standardization  159 P.E. James Dean Vogel

12.1 ­Introduction  159 12.2 ­Alphabet Soup  159 12.3 ­History  161 12.4 ­Compare and Contrast  161 12.5 ­Collaboration and Alignment Lead to Standardization  162 12.6 ­General SUT Efforts  163 12.7 ­Leachables and Extractables  164 12.8 ­Particulates in SUT  164 12.9 ­Change Notification  165 12.10 ­SUT System Integrity  165 12.11 ­SUT User Requirements  165 12.12 ­Connectors  165 12.13 ­SUT Design Verification  165 12.14 ­Summary and Conclusions  166 Nomenclature  166 References  166 Further Reading  166 13

Environmental Impacts of Single‐Use Systems  169 William G. Whitford, Mark A. Petrich, and William P. Flanagan

13.1 ­Introduction  169 13.2 ­Sustainability  169 13.3 ­The Evolution of SU Technologies  169 13.4 ­Implications in Sustainability  172 13.5 ­LCA – A Holistic Methodology  172 13.6 ­LCA Applied to SU Technologies  173 13.6.1 Early Attempts to Examine the Environmental Aspects of SU Technologies  173 13.6.2 LCA Applied to SU Technologies  173 13.7 ­Sustainability Efforts in the BioPharma Industry  175

Contents

13.8 ­End‐of‐Life (Waste) Management  177 13.9 ­Summary and Conclusions  178 Nomenclature  178 ­References  178 14

Design Considerations Towards an Intensified Single‐Use Facility  181 Gerben Zijlstra, Kai Touw, Michael Koch, and Miriam Monge

14.1 ­Introduction  181 14.2 ­Moving Towards Intensified and Continuous Processing  181 14.3 ­Methodologies for Continuous and Intensified Single‐Use Bioprocessing  183 14.4 ­Process Development for Intensified Biomanufacturing Facilities  184 14.5 ­The Intensified Biomanufacturing Facility  184 14.6 ­Process Automation for Commercial Manufacturing Facilities  187 14.7 ­Intensified Upstream Processing  187 14.8 ­Intensified Downstream Processing  189 14.9 ­Summary and Conclusions  191 Acknowledgments  191 Nomenclature  191 ­References  191 15

Single‐Use Technologies in Biopharmaceutical Manufacturing: A 10‐Year Review of Trends and the Future  193 Ronald A. Rader and Eric S. Langer

15.1 ­Introduction  193 15.2 ­Background  193 15.3 ­Methods  194 15.4 ­Results  194 15.4.1 Market (Facilities) Distribution  194 15.4.2 Single‐Use Systems Market Estimates  195 15.4.3 Market Trends and Perceptions  196 15.5 ­Discussion  197 15.6 ­Conclusions  199 Nomenclature  200 ­References  200 Part II  16

Application Reports and Case Studies  201

Single‐Use Process Platforms for Responsive and Cost‐Effective Manufacturing  203 Priyanka Gupta, Miriam Monge, Amelie Boulais, Nitin Chopra, and Nick Hutchinson

16.1 ­Introduction  203 16.2 ­Standardized Single‐Use Process Platforms for Biomanufacturing  204 16.3 ­Implementing Single‐Use Process Platforms  204 16.4 ­Economic Analysis Comparing Stainless Steel with Single‐Use Process Platforms  207 16.5 ­Summary and Conclusions  209 Nomenclature  209 ­References  210 17

Considerations on Performing Quality Risk Analysis for Production Processes with Single‐Use Systems  211 Ina Pahl, Armin Hauk, Lydia Schosser, and Sonja von Orlikowski

17.1 ­Introduction  211 17.2 ­Quality Risk Assessment  211 17.3 ­Terminology and Features  212 17.4 ­Current Industrial Approach for Leachable Assessment in Biopharmaceutical Processes  212

xi

xii

Contents

17.5 ­Holistic Approach to Predict Leachables for Quality Risk Assessment  214 17.6 ­Summary and Conclusions  215 Nomenclature  217 ­References  217 18

How to Assure Robustness, Sterility, and Performance of Single‐Use Systems: A Quality Approach from the Manufacturer’s Perspective  219 Simone Biel and Sara Bell

18.1 ­Introduction  219 18.2 ­Component Qualification  219 18.3 ­Validation of Product Design  220 18.3.1 Sterility Validation and Quarterly Dose Audit Approach  220 18.3.1.1 Initial Validation of Gamma Sterilization  220 18.3.1.2 Quarterly Dose Audits  221 18.3.1.3 Dose Maps  221 18.3.2 Integrity Assurance  222 18.3.2.1 Manufacturers’ Integrity Testing  222 18.3.2.2 Packaging and Shipping Validation  222 18.3.2.3 Point‐of‐Use Integrity Testing  223 18.3.3 Stability Studies  223 18.4 ­Manufacturing and Control  224 18.4.1 Process Flow  224 18.4.2 Cleanroom Classification  224 18.4.3 Validation of Single‐Use System Manufacturing Process  224 18.5 ­Operator Training, Performance Culture  225 18.5.1 In‐Process Controls  225 18.5.2 Quality Control: Release Tests  225 18.6 ­Particulate Risk Mitigation  225 18.7 ­Change Management  225 18.7.1 Levels of Change  226 18.7.2 Qualification  226 18.8 ­Summary and Conclusions  226 Nomenclature  227 ­References  227 19

How to Design and Qualify an Improved Film for Storage and Bioreactor Bags  229 Lucie Delaunay, Elke Jurkiewicz, Gerhard Greller, and Magali Barbaroux

19.1 ­Introduction  229 19.2 ­Materials, Process, and Suppliers Selection  229 19.3 ­Biological Properties  229 19.4 ­Specifications and Process Design Space  231 19.5 ­Process Control Strategy  233 19.6 ­Summary and Conclusions  233 Nomenclature  233 ­References  233 20

An Approach for Rapid Manufacture and Qualification of a Single‐Use Bioreactor Prototype  235 Stephan C. Kaiser

20.1 ­Introduction  235 20.2 ­About the Development Process of a Single‐Use Bioreactor  235 20.2.1 Conceptual Design and Software Tools  235 20.2.2 Molding vs. Rapid Prototyping  236 20.2.3 Engineering Characterization, Sterilization, and Qualification  239 20.3 ­Summary and Conclusions  243

Contents

­

Nomenclature  244 ­References  244 21

Single‐Use Bioreactor Platform for Microbial Fermentation  247 Parrish M. Galliher, Patrick Guertin, Ken Clapp, Colin Tuohey, Rick Damren, Yasser Kehail, Vincent Colombie, and Andreas Castan

21.1 ­Introduction  247 21.2 ­General Design Basis for Microbial SUFs  247 21.3 ­SUF Design Criteria and Approach – Heat Transfer  247 21.3.1 Engineering Principles – Total Heat Load and Transfer  247 21.3.2 Approach to Design of Heat Transfer/Removal Features for SUFs  248 21.4 ­SUF Design Criteria and Approach – Oxygen Transfer  249 21.4.1 Engineering Principles for Total Oxygen Demand and Transfer  249 21.4.2 Approach to Design Oxygen Transfer Features for SUFs  249 21.4.2.1 Vessel Pressure  249 21.4.2.2 Oxygen Enrichment of Sparge Gas  249 21.4.2.3 Microporous Spargers Versus Drilled Hole Spargers  250 21.4.2.4 Power of the Agitation System to Supply High OTR  250 21.4.2.5 Calculating Agitation Power Required to Meet the Maximum OUR  250 21.5 ­SUF Design Criteria and Approach – Mixing  251 21.5.1 Engineering Principles of Mixing  251 21.5.2 Modeling and Empirical Measurements of Mixing Time  251 21.5.3 Effect of Different Impeller Types  251 21.6 ­Operational Considerations for SUFs  252 21.6.1 Liquid Management  252 21.6.2 Media Sterilization  252 21.7 ­Case Studies  252 21.7.1 Heat Transfer Tests of SUFs (50 and 500 l)  252 21.7.2 Oxygen Mass Transfer Tests of SUFs (50 and 500 l)  252 21.7.3 Escherichia coli Fermentation in a 50 l SUF  252 21.7.4 Yeast Fermentation in a 50 l SUF  253 21.7.5 Pseudomonas fluorescens Fermentation in a 50 l SUF  254 21.7.6 Fermentation of E. coli in a 500 l SUF  254 21.7.7 Fermentation of Haemophilus Influenzae in a 500 l SUF  254 21.8 ­Summary and Conclusions  256 Nomenclature  257 ­References  258 22

Engineering Parameters in Single‐Use Bioreactors: Flow, Mixing, Aeration, and Suspension  259 Martina Micheletti and Andrea Ducci

22.1 ­Introduction  259 22.2 ­Stirred Bioreactors  259 22.2.1 Flow Dynamics in Mobius CellReady  260 22.2.2 Mixing Dynamics in Mobius CellReady  260 22.3 ­Orbitally Shaken Bioreactors  262 22.3.1 Flow Dynamics  262 22.3.2 Aeration – Interfacial Area  263 22.3.3 Mixing Dynamics  265 22.3.4 Suspension Dynamics  265 22.4 ­Rocking Bag  267 22.5 ­Summary and Conclusions  268 Nomenclature  268 ­References  268

xiii

xiv

Contents

23

Alluvial Filtration: An Effective and Economical Solution for Midstream Application (e.g. Cell and Host Cell Protein Removal)  271 Ralph Daumke, Vasily Medvedev, Tiago Albano, and Fabien Rousset

23.1 ­Introduction  271 23.1.1 Alluvial Filtration  271 23.1.2 Diatomaceous Earth Filtration  271 23.1.3 Depth Filtration  272 23.1.4 DAISEP MabXpure Technology  272 23.2 ­Case Study 1: Cell Removal  272 23.2.1 Background  272 23.2.2 Materials and Methods  272 23.2.2.1 Reagents and Equipment  272 23.2.2.2 Experimental Methods  272 23.2.2.3 Analytical Methods  273 23.2.3 Results and Discussion  273 23.2.3.1 Results 273 23.2.3.2 Discussion 273 23.3 ­Case Study 2: HCP Removal  275 23.3.1 Background  275 23.3.2 Materials and Methods  275 23.3.3 Procedure  275 23.3.3.1 Static Mode  275 23.3.3.2 Dynamic Mode  275 23.3.4 Results  275 23.3.5 Discussion  276 23.4 ­Summary and Conclusions  276 Nomenclature  277 ­References  277 24

Single‐Use Continuous Downstream Processing for Biopharmaceutical Products  279 Marc Bisschops, Britta Manser, and Martin Glenz

24.1 ­Introduction  279 24.2 ­Continuous Multicolumn Chromatography  279 24.3 ­Single‐Use Continuous Downstream Processing  280 24.3.1 Continuous Chromatography for Fed‐Batch Processes  280 24.3.2 Continuous Chromatography for Perfusion Processes  282 24.4 ­Summary and Conclusions  283 ­References  283 25

Single‐Use Technology for Formulation and Filling Applications  285 Christophe Pierlot, Alain Vanhecke, Kevin Thompson, Rainer Gloeckler, and Daniel Kehl

25.1 ­Introduction  285 25.2 ­Challenges in Formulation and Filling  285 25.3 ­End‐User Requirements  286 25.4 ­Quality by Design  287 25.5 ­Hardware Design and Usability  288 25.6 ­Single‐Use Technology, Arrangement, and Operation  290 25.7 ­Summary and Conclusions  293 Nomenclature  294 ­References  294 26

Facility Design Considerations for Mammalian Cell Culture  295 Sue Walker

26.1 ­Introduction  295 26.2 ­Generic Case Study  295

Contents

26.2.1 Generation of the Process Model  295 26.2.2 Generation of the Facility Description  297 26.2.2.1 Manufacturing Areas  299 26.2.2.2 Support Areas  300 26.3 ­Summary and Conclusions  301 Nomenclature  301 ­References  301 27

Progress in the Development of Single‐Use Solutions in Antibody–Drug Conjugate (ADC) Manufacturing  303 Diego R. Schmidhalter, Stephan Elzner, and Romeo Schmid

27.1 ­Introduction  303 27.2 ­Challenges for the Use of Disposables in ADC Processes  304 27.2.1 Use of Organic Solvents  304 27.2.2 Safety and Handling of HPAPIs  305 27.3 ­Key Unit Operations  306 27.3.1 Reactions in Stirred Tanks  306 27.3.2 Tangential Flow Filtration  306 27.3.3 Chromatography  306 27.3.4 Filtration and Transfers  306 27.3.5 Bulk Drug Substance Freeze and Thaw  306 27.4 ­Cysteine Conjugation Process – An ADC Production Process Case Study  308 27.5 ­Summary and Conclusions  309 Acknowledgment  309 Nomenclature  309 ­References  310 28

Single‐Use Processing as a Safe and Convenient Way to Develop and Manufacture Moss‐Derived Biopharmaceuticals  311 Holger Niederkrüger, Andreas Busch, Paulina Dabrowska‐Schlepp, Nicola Krieghoff, Andreas Schaaf, and Thomas Frischmuth

28.1 ­Introduction  311 28.2 ­Case Study  311 28.2.1 Introduction to Greenovation Biotech’s Bryotechnology  311 28.2.2 Frame of Case Study  312 28.2.3 Process Description  312 28.2.3.1 Moss Cell Line  312 28.2.3.2 Cell Banking  312 28.2.3.3 USP and DSP  313 28.2.3.4 Moss Metabolite Study  315 28.2.3.5 Development of a Moss‐Specific HCP Assay  315 28.2.3.6 L&E Study for Illuminated Bag Films  315 28.2.4 Assessment of Case Study  317 28.3 ­Summary and Outlook  317 Nomenclature  317 ­References  318 29

Single‐Use Technologies Used in Cell and Gene Therapy Manufacturing Need to Fulfill Higher and Novel Requirements: How Can this Challenge Be Addressed?  319 Alain Pralong and Angélique Palumbo

29.1 ­Introduction  319 29.2 ­Promise of Cell and Gene Therapy  320 29.2.1 Cancer – A Long‐Known Disease  320 29.2.2 Evolution of Cancer Treatment  321 29.2.3 The Process of Adoptive T‐Cell Therapy  322

xv

xvi

Contents

29.2.3.1 Manufacturing of Autologous Engineered Adoptive T‐Cell Therapeutics  322 29.2.3.2 Manufacturing of Allogeneic Engineered Adoptive T‐Cell Therapeutics  322 29.3 ­Considerations for Biopharmaceutical Industry and Conclusion  322 Nomenclature  325 ­References  325 30

Single‐Use Bioreactors for Manufacturing of Immune Cell Therapeutics  327 Ralf Pörtner, Christian Sebald, Shreemanta K. Parida, and Hans Hoffmeister

30.1 ­Introduction  327 30.2 ­The Particular Nature of Immune Cell Therapeutics  327 30.3 ­Uncertain Mass Production of Immune Cells for Therapy  328 30.4 ­Technical Standards Required for Immune Cell ATMP Manufacturing  329 30.5 ­Techniques for Expansion of Immune Cells  329 30.6 ­Case Study ZRP System Consisting of GMP Breeder, Control Unit, and Software  330 30.7 ­Summary and Conclusions  330 Nomenclature  332 ­References  332 Index  335

xvii

List of Contributors Tiago Albano

Katharina Dahlmann

Univercells SA, Gosselies, Belgium

Institut für Technische Chemie, Gottfried Wilhelm Leibniz Universität Hannover, Germany

Magali Barbaroux

Sartorius Stedim FMT S.A.S., Aubagne, France Sara Bell

MilliporeSigma, Bedford, MA, USA Simone Biel

Rick Damren

GE Healthcare Life Sciences, Marlborough, MA, USA Ralph Daumke

FILTROX AG, St. Gallen, Switzerland

Merck Chemicals GmbH, an affiliate of Merck KGaA, Darmstadt, Germany

Lucie Delaunay

Marc Bisschops

Andrea Ducci

Pall International Sàrl, Fribourg, Switzerland

University College London, London, UK

Amelie Boulais

Dieter Eibl

Sartorius Stedim Biotech GmbH, Göttingen, Germany

School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology, Zurich University of Applied Sciences, Wädenswil, Switzerland

Andreas Busch

Greenovation Biotech GmbH, Freiburg im Breisgau, Germany

Sartorius Stedim FMT S.A.S., Aubagne, France

Regine Eibl

Meissner Filtration Products, Camarillo, CA, USA

School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology, Zurich University of Applied Sciences, Wädenswil, Switzerland

Andreas Castan

Stephan Elzner

GE Healthcare Life Sciences, Uppsala, Sweden

Lonza AG, Visp, Switzerland

Nitin Chopra

William P. Flanagan

Sartorius Stedim Biotech GmbH, Göttingen, Germany

Aspire Sustainability, Albany, NY, USA

Ken Clapp

Thomas Frischmuth

GE Healthcare Life Sciences, Marlborough, MA, USA

Greenovation Biotech GmbH, Freiburg im Breisgau, Germany

Chuck Capron

Vincent Colombie

Sanofi Pasteur, Campus Merieux, Marcy l’Etoile, France Paulina Dabrowska‐Schlepp

Greenovation Biotech GmbH, Freiburg im Breisgau, Germany

Parrish M. Galliher

GE Healthcare Life Sciences, Marlborough, MA, USA Martin Glenz

Pall International Sàrl, Fribourg, Switzerland

xviii

List of Contributors

Rainer Gloeckler

Matthias Kraume

Swissfillon AG, Visp, Switzerland

Technische Universität Berlin, Fakultät III Verfahrenstechnik, Lehrstuhl Verfahrenstechnik, Berlin, Germany

Gerhard Greller

Sartorius Stedim Biotech GmbH, Göttingen, Germany Patrick Guertin

Nicola Krieghoff

GE Healthcare Life Sciences, Marlborough, MA, USA

Greenovation Biotech GmbH, Freiburg im Breisgau, Germany

Priyanka Gupta

Eric S. Langer

Sartorius Stedim Biotech GmbH, Göttingen, Germany Armin Hauk

Sartorius Stedim Biotech GmbH, Göttingen, Germany Hans Hoffmeister

Zellwerk GmbH, Oberkrämer, Germany Nick Hutchinson

Sartorius Stedim Biotech GmbH, Göttingen, Germany Ernest Jenness

MilliporeSigma, Bedford, MA, USA Valentin Jossen

School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology, Zurich University of Applied Sciences, Wädenswil, Switzerland Christian Julien

Meissner Filtration Products, Camarillo, CA, USA Elke Jurkiewicz

BioPlan Associates, Inc., Rockville, MD, USA Britta Manser

Pall International Sàrl, Fribourg, Switzerland Vasily Medvedev

Univercells SA, Gosselies, Belgium Martina Micheletti

University College London, London, UK Miriam Monge

Sartorius Stedim Biotech GmbH, Göttingen, Germany Holger Niederkrüger

Greenovation Biotech GmbH, Freiburg im Breisgau, Germany Ina Pahl

Sartorius Stedim Biotech GmbH, Göttingen, Germany Barbara A. Paldus

Sekhmet Ventures, Portola Valley, CA, USA

Sartorius Stedim Biotech GmbH, Göttingen, Germany

Angélique Palumbo

Stephan C. Kaiser

Shreemanta K. Parida

Thermo Fisher Scientific, Santa Clara, CA, USA Frank Lammers

Sanofi, Frankfurt am Main, Germany Yasser Kehail

GE Healthcare Life Sciences, Marlborough, MA, USA Daniel Kehl

Swissfillon AG, Visp, Switzerland Michael Koch

Sartorius Stedim Biotech GmbH, Göttingen, Germany

Science & Tech ENABLE GmbH, Solothurn, Switzerland Zellwerk GmbH, Oberkrämer, Germany Mark A. Petrich

Merck & Co., Inc., West Point, PA, USA Christophe Pierlot

Pall Biotech, Hoegaarden, Belgium Catherine Piton

Pall Biotech, Saint‐Germain‐en‐Laye, France Alain Pralong

Pharma‐Consulting ENABLE GmbH, Solothurn, Switzerland

List of Contributors

Ralf Pörtner

Dörte Solle

Institute of Bioprocess and Biosystems Engineering, Hamburg University of Technology, Hamburg, Germany

Institut für Technische Chemie, Gottfried Wilhelm Leibniz Universität Hannover, Germany

Ronald A. Rader

BioPlan Associates, Inc., Rockville, MD, USA

Institut für Technische Chemie, Gottfried Wilhelm Leibniz Universität Hannover, Germany

Kenneth F. Reardon

Kevin Thompson

Department of Chemical and Biological Engineering, Colorado State University, Fort Collins, CO, USA

Pall Biotech, Portsmouth, UK

Sebastian Rothe

Sartorius Stedim Biotech GmbH, Göttingen, Germany

GE Healthcare Europe GmbH, Freiburg, Germany Fabien Rousset

DAICEL Bioseparations – Chiral Technologies Europe SAS, Illkirch, France Jens Rumsfeld

Sartorius Stedim Biotech GmbH, Göttingen, Germany Andreas Schaaf

Tobias Steinwedel

Kai Touw

Colin Tuohey

GE Healthcare Life Sciences, Marlborough, MA, USA Steven Vanhamel

Pall Biotech, Port Washington, NY, USA Alain Vanhecke

Pall Biotech, Hoegaarden, Belgium

Greenovation Biotech GmbH, Freiburg im Breisgau, Germany

The BioProcess Institute, North Kingstown, RI, USA

Thomas Scheper

Sonja von Orlikowski

Institut für Technische Chemie, Gottfried Wilhelm Leibniz Universität Hannover, Germany

Boehringer Ingelheim Pharma GmbH & Co. KG, Ingelheim, Germany

Cedric Schirmer

Sue Walker

School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology, Zurich University of Applied Sciences, Wädenswil, Switzerland

Engineering Consultant, Portsmouth, NH, USA

Romeo Schmid

Lonza AG, Visp, Switzerland

P.E. James Dean Vogel

Christian Weiner

Octapharma Pharmazeutika Produktionsges.m.b.H, Vienna, Austria Sören Werner

Lonza AG, Visp, Switzerland

School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology, Zurich University of Applied Sciences, Wädenswil, Switzerland

Lydia Schosser

William G. Whitford

Boehringer Ingelheim Pharma GmbH & Co. KG, Ingelheim, Germany

GE Healthcare, South Logan, UT, USA

Christian Sebald

Sartorius Stedim Biotech GmbH, Göttingen, Germany

Diego R. Schmidhalter

Zellwerk GmbH, Oberkrämer, Germany Alicja Sobańtka

Octapharma Pharmazeutika Produktionsges.m.b.H, Vienna, Austria

Gerben Zijlstra

xix

xxi

Preface Single‐use devices have become a major part of the ­biopharmaceutical production process. They now make up 85% of the equipment in preclinical bioprocessing and are increasingly being employed in the commercial manufacture of biopharmaceuticals. It is in upstream processing, which can be accomplished entirely with ­single‐use technology, where they are used with greatest diversity, for example, in the manufacturing of modern antibodies and vaccines. Single‐use solutions are also, however, available for downstream processing and for Fill & Finish which are accepted by users. Today, the first fully single‐use production facilities have already become a reality. It seems that users have more confidence in single‐use technology, which can be explained by the further development and the improved design of such devices. The new generations of single‐use devices are more robust and ­easier to handle than their predecessors. Possible problems, such as leakage and integrity, have already been addressed by the suppliers during the manufacturing process. Moreover, progress has been made in film technologies, bioreactor design, sensor techniques, and automation. The second edition of the book Single‐Use Techno­ logy in Biopharmaceutical Manufacture consists of an introduction section for beginners and a case‐study

collection for advanced‐level readers. It summarizes the latest developments in single‐use technologies. In  addition to a presentation of single‐use systems as applied to different unit operations and to platform technologies, their selection, implementation, and level of trouble‐free usage are discussed. This includes approaches to intensify bioprocesses and to  realize continuous processes but also to aspects of quality assurance and standardization, the influence of single‐ use technology on the environment, and the importance of risk analysis. We would like to thank all authors for their valuable contributions to the new edition of this book. We would also like to extend our special thanks to the management of the Department for Life Sciences and Facility Management of the Zurich University of Applied Sciences for their support in realizing this book. We hope that the new edition of Single‐Use Technology in Biopharmaceutical Manufacture will be helpful for bachelor and master students of biotechnology and related fields, for experienced practitioners who are developing as well as producing biopharmaceuticals and designing production facilities, and, finally, for those who intend to begin using disposables. Regine and Dieter Eibl

1

Part I Basics

3

1 Single‐Use Equipment in Biopharmaceutical Manufacture A Brief Introduction Dieter Eibl and Regine Eibl School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology, Zurich University of Applied Sciences, Wädenswil, Switzerland

1.1 ­Background The term “biopharmaceutical” was first used in the early 1980s [1] when recombinant, commercially manufac­ tured insulin, a therapeutic protein for diabetes patients, was introduced. In the United States and Europe, the most frequently used definition is that of a pharmaceuti­ cal manufactured by biotechnological methods with organisms, or their functional components, which have a biological origin. Following this definition, all recombi­ nant proteins, monoclonal antibodies (mAbs), vaccines, blood/plasma‐derived products, nonrecombinant cul­ ture‐derived proteins, and cultured cells, in addition to tissues from human or animal origin and nucleic acids, are considered to be biopharmaceuticals [2–4]. The majority of the above are classified as biologicals (or ­biologics) by regulatory agencies [5]. Traditional phar­ maceutical products, such as chemical compounds extracted from plants, secondary metabolites from microbial and plant cell cultures, and synthetic peptides, which may not comply with the above definition, are  more often regarded as non‐biopharmaceuticals. Irrespective of differences in definition, recombinant protein pharmaceuticals constitute an important cate­ gory of biopharmaceuticals. The most significant protein pharmaceuticals available include hormones such as erythropoietin, enzymes such as the human plasminogen activator, vaccines such as Flucelvax, and mAbs such as bevacizumab. It is worth mentioning that the top 10 best‐selling drugs are domi­ nated by therapeutic mAbs today [5]. In most cases, protein pharmaceuticals are produced with mammalian cell lines. During the last few years, Chinese hamster ovary cell lines have increasingly dis­ placed earlier mammalian cell production systems such as hybridomas or embryonic feline lung fibroblast cell lines [6, 7]. Further production organisms of choice for

protein pharmaceuticals are microbial cells [8] (see also Chapter  21), plant cells [9] (see also Chapter  28), and insect cells cultivated in conjunction with the bac­ ulovirus expression vector system [10]. The worldwide demand for protein pharmaceuticals (and, in particular, protein therapeutics) has resulted in increased efforts to expand the process efficiency over the past 10 years. It is undoubtedly the case that the huge growth in knowledge in molecular and cell biology has led to high‐productivity cell lines and improved culture media. These cell lines provide product titers exceeding 3 g/l in fed batch mode and contribute to shrinking bio­ reactor size, which is associated with cost savings [11]. Further cost savings can be achieved by replacing stain­ less steel with single‐use equipment in the production process [12, 13]. The present chapter introduces the reader to the area of single‐use technology. In addition to terminology, advantages and disadvantages of existing single‐use devices will be described. Based on a schematic of a ­typical production process for a protein therapeutic, an overview of currently available single‐use devices and a categorization approach will be presented. Moreover, the main criteria for implementing single‐use systems in biopharmaceutical production processes are summa­ rized, and current concepts concerning single‐use ­production facilities are briefly explained.

1.2 ­Terminology and Features As the term “single‐use” (or “disposable”) implies, such systems are only ever used once. Disposables currently in use originated in the fields of medical care (e.g. rubber gloves, sterile swabs, and the technology for intravenous applications) and infant care (e.g. paper towels and disposable diapers). With the exception of special ­

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

4

1  Single‐Use Equipment in Biopharmaceutical Manufacture

­rotective clothing and consumables (e.g. swabs and p paper towels), single‐use products are typically fabri­ cated from plastics approved by the Food and Drug Administration (see also Chapter 8), such as polyethylene, polystyrene, polytetrafluoroethylene, polypropylene, or ethylene vinyl acetate. These materials are typically sup­ plemented with additives to aid performance and/or prolong usable life [14, 15], thereby ensuring their suita­ bility in biopharmaceutical manufacturing applications. In all cases, the product contact surfaces are free of ­animal‐derived components. Disposables can be rigid (molded systems) or flexible (bags made from multilayer films) and are often supplied presterilized, having been gamma irradiated at dose lev­ els between 25 and 50 kGy [16, 17], although some are autoclaved or sterilized with gas. This eliminates the need for subsequent sterilization of the equipment, such as the steam sterilization normally required for stainless‐ steel components. Disposables can, therefore, quickly be brought into operation. On completion of the process operations, the disposables used are decontaminated and discarded. Thus, time‐consuming and expensive cleaning procedures which may require the use of cor­ rosive chemicals (which could potentially pose a health hazard to the operator) and water‐for‐injection, often considered as a bottleneck in traditional biopharmaceu­ tical facilities, are no longer required. Disposable technology is often regarded as greener, due to the reduced requirements for cleaning and sterili­ zation (see also Chapter  13). Furthermore, equipment turnaround time is reduced, and process and product changes can be more easily accommodated (a particular advantage in the manufacture of multiple products) when neither cleaning nor sterilization is required [17]. Similarly, the potential for product cross contamination and microbial contamination is reduced, and the require­ ments for validation and in‐process documentation are minimized [15, 18, 19]. Further benefits of disposables include savings in time (e.g. development time, manufac­ turing time, and time to market), cost reductions (e.g. capital investment and cost of goods sold), and a reduc­ tion of the facility’s footprint. It can be concluded that disposables may offer distinct advantages compared to their reusable counterparts when selected and used cor­ rectly. To summarize, they can be smaller, safer, greener, faster, and more flexible, while offering savings both in terms of capital outlay and operating costs (see Table 1.1). Yet, there are still limitations to the use of disposables due to the chemical, biological, and physical properties of the plastic material. Besides leakage (see also Chapter 2), the primary risk associated with the use of disposables is the potential migration of undesired com­ ponents from the plastic material (see also Chapters 8, 11, 17, and 18). Main undesired contaminants may either

Table 1.1  Summary of advantages and limitations of single‐use equipment. Pros

Safer: High bio- and process safety ●● Sterilized ●● Preassembled ●● Decreased risk of microbial contamination and cross contamination ●● Facilitates qualification and validation Greener ●● Reduced requirements for cleaning and sterilization Faster and more flexible ●● Easier process and product change Cheaper: Saving of time and cost ●● Reduction of cultivation, cleaning, sterilization, qualification, and maintenance requirements ●● Lower capital investment, reduced infrastructure and maintenance costs Smaller ●● Reduced facility footprint

Cons

Material properties Breakage and leakage ●● Leachables and extractables, particulates Scalability ●● Limited by the properties and fabrication of the polymer materials Running costs and wastes ●● Increased operating costs (costs of solid waste disposal and consumables) ●● Ongoing replacement of the disposables Automation level, sensors ●● No high‐level automation solutions ●● Restricted availability of disposable sensors Lack of standardization ●● Supplier dependence Training of staff ●● Increasing requirement with rising culture volume ●●

be leachables (which may migrate under process condi­ tions over time) or extractables (which may migrate when exposed to aggressive process conditions such as high temperatures) [20, 21]. Another topic that has been raised by the increasing implementation of single‐use ­ articulate devices for final Fill & Finish is the detection of p contamination over the past years (see also Chapter 18) [22, 23]. Additional issues which limit the use of dispos­ ables are restricted scalability (due to the mechanical strength of the material), the limited availability of sin­ gle‐use sensors (see also Chapter  6), and the lack of advanced automation techniques (see also Chapter 7). So far, the replacement of disposable components con­ stitutes an increase in operating costs and contributes to the increased cost of solid waste disposal and consum­ ables. Another weakness of single‐use systems is the dependence on suppliers (see also Section 1.4) resulting from lack of standardization (see also Chapter  13). Furthermore, it is worth noting that extra training of staff may be necessary as the scale of a manufacturing facility incorporating disposables increases. A challenge that should also not be underestimated is the packaging of single‐use systems which covers the system integrity at the supplier level as well as in manufacturing and the maintenance of sterility. Thus, a thorough investigation

1.3  Single‐Use Systems in Production Processes for Therapeutic Proteins such as mAbs: Product Overview and Classification

is recommended to determine whether the benefits of disposable systems are sufficient to overcome their dis­ advantages in any particular manufacturing scenario.

1.3 ­Single‐Use Systems in Production Processes for Therapeutic Proteins such as mAbs: Product Overview and Classification As illustrated in Figure  1.1, a typical process for the ­manufacture of a drug product (DP) such as therapeutic proteins (mAbs) includes four main processing stages: (i) upstream processing, (ii) downstream processing, (iii) final formulation and filling, and (iv) labeling and pack­ aging. In the upstream processing stage, culture media and buffers are prepared (mixed, sterilized by 0.1 and 0.2 μm filters, stored, and transported), seed and inocu­ lum train are produced, and a so‐called active pharma­ ceutical ingredient (API) is expressed in the production bioreactor (see Figure 1.1). The API which is, with only a few exceptions such as membrane proteins, normally secreted into the culture broth has to be separated from cells and clarified after harvesting. The subsequent downstream procedures [24–30] (see also Chapters 9, Manufacturing of the API

Pharmaceutical manufacturing Final formulation and filling

Upstream processing Media preparation

10, 23, and 24), which produce a drug substance (DS), ensure the reduction of product impurities (e.g. protein A, host cell proteins, desoxyribonucleic acid, and aggre­ gates) to an acceptably low level and include virus clear­ ing (inactivation and removal by filtration). Consequently, the API must be further concentrated, separated, and purified, requiring chromatography processes (affinity chromatography, anion‐exchange chromatography, ­cation‐exchange chromatography, and hydrophobic inter­ action chromatography) and crossflow filtration (ultra and diafiltrations). Liquid storage and transportation, and buffer preparation also form part of the downstream processing stage. The liquid DS solution is formulated through the addition of stabilizers prior to being steri­ lized by filtration and/or aseptically poured into sterile containers. The DS may also be stored or transported when it is deep frozen prior to the Fill & Finish opera­ tions. The DS is then labeled and packaged to become the commercially available DP. Nowadays, the developer and manufacturer of a thera­ peutic protein can choose among a multitude of single‐ use devices from different suppliers for all stages of the production process. Figure 1.2 provides an overview of the primary disposables currently utilizable in therapeu­ tic protein manufacturing. Single‐use devices can be classified into three groups: expendable laboratory items, simple peripheral elements

Inoculum production and cell cultivation

Formulation

Fermentation Cell separation and clarification Downstream processing

Compounding

Labeling and packaging Labeling

Buffer preparation

Packaging

Concentration and capturing Polishing (virus clearing included)

Finished drug product

DS

Figure 1.1  Schematic of a typical manufacturing process for therapeutic proteins such as mAbs.

DP

5

6

1  Single‐Use Equipment in Biopharmaceutical Manufacture

Single-use

• • • • • • • • • •

Expendable laboratory items

Simple peripheral elements

Analyzer sample caps Culture containers Flasks Microtiter plates Petri dishes Pipette and pipette tips Protective clothing Syringes Test and centrifuge tubes Vent and liquid filters

• Aseptic transfer systems • 2D-, 3D-bags, bag manifold systems, bag handling systems • Connectors, tri-clamps • Flexible tubing • Fittings, molded fittings • Liquid containment bags • Stopper, closure containers, protective caps • Tank liners • Valves

Equipment for unit operations and platform technologies

• • • • • • • • •

Bioprocess containers Bioreactors Centrifuges Chromatography systems Depth filter systems Freeze–thaw-systems Isolators Membrane adsorbers Micro-, ultra-, diafiltrationdevices

• Mixing systems • Pumps

Figure 1.2  Primary categories of disposables utilizable for the development and manufacture of therapeutic proteins [31]. Source: Reproduced with permission of John Wiley & Sons.

(stand‐alone components), and multi‐component ­systems for unit operations and platform technologies. Thanks to single‐use bioreactors (Chapter  4) together with bags for storage as well as transportation (Chapter 2), single‐use mixers (Chapter  3), single‐use plastic hoses, single‐use plastic fittings, single‐use connectors and sampling systems, and single‐use pumps (Chapter  5), upstream processing carried out entirely with single‐use technology up to the mid‐volume scale has become pos­ sible. Leak test systems (Chapter 2) and novel connectors (multi‐utilizable, hybrid, and neutral versions) have additionally improved safety in both upstream and downstream processing. Single‐use systems preferably applied in downstream processing (Chapters 9 and 10) include those for centrifugation and filtration (micro‐, ultra‐, and tangential flow filtration), when biomass has to be separated, culture broth has to be clarified, or a virus has to be separated or inactivated. In addition, ­single‐use membrane adsorbers and prepacked single‐ use chromatography systems have become increasingly common. Finally, the formulation and filling process steps are already able to be executed with single‐use sys­ tems such as single‐use storage systems, single‐use fil­ ters, single‐use mixers, single‐use isolators, single‐use dosage systems, single‐use needles, etc. (Chapter 25). Each and every key player in single‐use technology now offers single‐use process platform technologies (for media preparation, inoculum production, fermentation and biomass separation, virus separation and virus inac­ tivation, formulation, and filling). Product examples

include the ReadyToProcess and the FlexFactory series (GE Healthcare), the Mobius series (Merck), the Allegro series (Pall), the FlexAct series (Sartorius Stedim Biotech), and the HyPerforma series (Thermo Scientific). These process platforms support the rational imple­ mentation of disposables and process intensification (Chapters 14 and 16). Disposables may be used in the same manner as their stainless‐steel counterparts, provided due consideration is given to their specific characteristics. The user’s requirements constitute the primary criteria in the ­decision‐making process, while the projected product demand and the optimized usage of the asset must also be taken into account. The performance of the dispos­ able, the associated costs, and the security of the supply chain must also be considered, while the risk of using a disposable must be minimized. Disposables pose a par­ ticular challenge in terms of assessing the technical risk associated with their use and the security of their supply chain. The majority of products have not been standard­ ized, and therefore the security of supply, outlined in Figure 1.3, is of paramount importance when consider­ ing the utilization of disposables [32]. The essential prerequisite for the implementation of disposables in biopharmaceutical manufacturing is a thorough under­ standing of the associated risks and the appropriate management thereof. As described by Pora and Rowlings [33], and Sinclair and Monge [34–37], numerous factors must be considered. A risk analysis has to be done as shown by Merseburger et al. [38, 39] or Merck [40].

1.5  Summary and Conclusions

Questions concerning supply > Will the product/technology selected still be available in 5 or 10 years? > Is the product/technology selected free of any intellectual property issues? > Is the supplier selected financially strong and does it have a proven track record? > Does the supplier selected have sufficient production capacity and facilities in case of problems or emergencies? > Is the supplier selected likely to stay independent or will this supplier be acquired in the foreseeable future?

Figure 1.3  Questions concerning security of supply.

1.4 ­Single‐Use Production Facilities As is clear from Figure  1.4, the possibility exists to ­manufacture therapeutic mAbs in a complete single‐use production facility as done, for example, by WuXi Pharmatech in Shanghai. However, fully single‐use pro­ duction facilities only constitute around 10% of all such facilities and, as such, are still an exception in the manu­ facturing of biopharmaceuticals [41]. Hybrid production facilities, in which single‐use and reusable devices are combined, are currently predominant. This is ascribed to the fact that downstream processing is not yet univer­ sally realizable with single‐use devices in every case depending on the scale and process mode (see also Chapters 9 and 24). A single‐use production facility (see also Chapter 26) exists when transport as well as storage of raw materials, intermediate and final products, and mass conversion take place mainly or entirely in single‐use devices. A dis­ tinction can be made between the single‐use facilities of the closed type (Figure 1.5a) and the single‐use facilities operating in stations (Figure 1.5b). A single‐use facility of the closed type (e.g. manifold system for buffer and media preparation) is characterized by the closed imple­ mentation of all process steps and is limited to small working volumes. The disposables are prefabricated and coupled in the sequence corresponding to the required process steps. The material is transported by using free drainage or pressure. In contrast, a single‐use facility operating in stations, where the material is moved from the process step to the next step by using transportable, single‐use containers, is more flexible [42, 43]. Single‐ use production facilities where upstream processing is executed entirely with single‐use devices, such as Shire’s facility for the manufacture of velaglucerase alfa in Lexington (Massachusetts), belong to single‐use facili­ ties operating in stations. Such facilities are regarded as

biofacilities of the future, allowing sustainable produc­ tions in smaller spaces and at lower investment costs. Their key feature is the flexible and modular design, for which four room concepts have gained acceptance: (i) the ballroom concept, (ii) the dance floor concept, (iii) the integration of modular cleanroom units into a ball­ room, and (iv) the assembly of modular, configurable units such as KUBio or FlexFactory modules (also referred to as blueprint facilities). The room concepts have been described in detail in Refs. [44, 45].

1.5 ­Summary and Conclusions The acceptance of single‐use systems, which are obtain­ able for all stages of the biopharmaceutical production process up to the mid‐volume scale, has increased in the production of biotherapeutics in the past 10 years (see also Chapter 15). This concerns, in particular, the devel­ opment and manufacture of mAbs, therapeutic hor­ mones, as well as enzymes and vaccines which are mainly produced with continuous animal and human cell lines. In fact, in upstream processing, single‐use technology is  a mainstream technology. Yet, although single‐use devices still have limitations and hybrid production facil­ ities are predominant, the first biotherapeutics and their biosimilars have already been manufactured in fully ­single‐use production facilities. It is expected that the number of implemented single‐use devices and single‐ use production facilities will continue to grow worldwide. There is a trend towards smaller, more modular, flexible, and sustainable biopharmaceutical productions in which the focus is on process intensification and continuous mode (see also Chapters 9, 14, 15, and 24). Perfusion is becoming increasingly popular (see also Chapter 14). In addition to biosimilars, antibody drug conjugates (see also Chapter 27) will contribute to the continuing growth

7

Buffer preparation

Media preparation

Inoculum production and cell cultivation BRX

Powder

Powder

BRX

BRX

BRX 1 ...10 l

MPB 500 l

BPB 500 l

API fermentation

API isolation and concentration FIL 5 m2 HAB

BRX 1 ...10 l

100 l

20 l

500 l

CHR

POB 400 I

50 I

600 l

FIL 1 l2.5 m2

MHB

BHB

25 cm FIL

FIL

BHB

Seed 2

Seed 1

MHB

API polishing FIL

POB 200 l

CHR

FIL

POB 200 l

0.5 m2/20 µm

POT 50 l

FIL 3 m2

Harvest bag

Chromatography 1 capture

DS formulation and filling

FIL

POB 50 l

Virus inactivation (pH shift)

FIL

BDS

BDS

16 l

16 l

Pool bag Clarification cooling/mixing

DS labelling and packaging

FIL

POB 50 l

Finished drug product

40 cm

40 cm Chromatography cation exchange 1 cycle

Virus filtration

Pool bag cooling/mixing/pH

Chromatography anion exchange 1 cycle Pool bag cooling/mixing

Ultra-/diafiltration

Pool bag cooling/mixing/pH

Formulation

Bulk filling station

Sterile filtration

Lead sheet API BDS BHB BPB

Depth filtration

BDS freezing and storing and thawing

FIL

POB 200 l

CHR

Production

BDS formulation and filling

FIL

1

Seed 3

Active pharmaceutical ingredient Bulk drug substance Buffer hold bag Buffer preparation bag

BRX CHR DS FIL HAB

Bioreactor Chromatography Drug substance Filtration Harvest bag

MHB MPB POB POT

Media hold bag Media preparation bag Pool bag Pool tank

Figure 1.4  Process flow diagram of a mid‐scale production for a mAb in which disposables are used.

Freezing/storing/thawing (FT 16)

Formulation

Vial filling station

Sterile filtration

­  References

(a)

(b) In-process-product in a disposable system

In-process-product in a disposable system

Process step

Process step in a new disposable system

Process step

Process step In-process-product in a new disposable system

Process step

Process step in a new disposable system

In-process-product in a new disposable system

In-process-product in a new disposable system

Material transport with pre-fabricated and coupled flexible tubing

Material transport with containers

Figure 1.5  Types of single‐use facilities: (a) single‐use facility of the closed type and (b) single‐use facility operating in stations.

of single‐use devices and their technology. Finally, ­single‐use technology will increasingly be requested in the future for the processing of primary cell lines such as

T‐cells and stem cells (see also Chapters 29 and 30), plant cells (see also Chapter 28), and microbial cells (see also Chapter 21).

Nomenclature API Active pharmaceutical ingredient BDS Bulk drug substance BHB Buffer hold bag BPB Buffer preparation bag BRX Bioreactor CHR Chromatography DP Drug product DS Drug substance

FIL Filtration HAT Harvest tank MHB Media hold bag MPB Media preparation bag mAbs Monoclonal antibodies POB Pool bag POT Pool tank

­References 1 Sinclair, A. (2009). An industry in transition. BioProcess

Int. 1 (Suppl.): 1. Walsh, G. (2002). Biopharmaceuticals and biotechnology 2 medicines: an issue of nomenclature. Eur. J. Pharm. Sci. 15: 135–138. Rader, R.A. (2005). What is a biopharmaceutical? Part 1: 3 (Bio) technology‐based definitions. BioExecutive Int. March: 60–65.

4 Rader, R.A. (2005). What is a biopharmaceutical? Part 2:

company and industry definitions. BioExecutive Int. May: 42–49. 5 Mathaes, R. and Mahler, H.C. (2018). Next generation biopharmaceuticals: product development. Adv. Biochem. Eng. Biotechnol. https://doi.org/10.1007/10_2016_57. Rader, R.A. (2008). Expression systems for process and 6 product improvement. BioProcess Int. 4 (Suppl.): 4–9.

9

10

1  Single‐Use Equipment in Biopharmaceutical Manufacture

7 Zhu, J. and Hatton, D. (2017). New mammalian

8

9

10

11

12

13

14

15 16

17

18

19

20

21

22

expression systems. Adv. Biochem. Eng. Biotechnol. https://doi.org/10.1007/10_2016_55. Jozala, A.F., Geraldes, D.C., Tundisi, L.L. et al. (2016). Biopharmaceuticals from microorganisms: from purification to purification. Braz J. Microbiol. 47 (Suppl. 1): 51–63. Santos, R.B., Abranches, R., Fischer, R. et al. (2016). Putting the spotlight back on plant suspension cultures. Front. Plant. Sci. 7: 297. https://doi.org/10.3389/ fpls.2016.00297. Siew, A. (2016). Large‐scale protein expression in baculovirus‐infected insect cells. BioPharm Int. 29 (12): 28–29. Kunert, R. and Reinhart, D. (2016). Advances in recombinant antibody manufacturing. Appl. Microbiol. Biotechnol. 100: 3451–3461. Lopes, A.G. and Brown, A. (2016). Practical Guide to Single‐use Technology: Design and Implementation. Shawbury, Shrewsbury, Shropshire: Smithers Rapra. Sandle T. (2018). Strategy for the adaption of single‐use technology. https://www. europeanpharmaceuticalreview.com/article/73939/ strategy‐for‐the‐adoption‐of‐single‐use‐technology (accessed 11 November 2018). Pora, H. and Rawlings, B. (2009). A user’s checklist for introducing single‐use components into process systems. BioProcess Int. 4: 9–16. Aranha, H. (2004). Disposable systems. BioProcess Int. 9 (Suppl.): 6–16. BSPA Subcommittee (2007). Guide to irradiation and sterilization validation of single‐use bioprocesses, part 1. BioProcess Int. 8: 32–39. BSPA Subcommittee (2007). Guide to irradiation and sterilization validation of single‐use bioprocesses, part 2. BioProcess Int. 9: 60–70. Hodge, G. (2004). Integrating emerging technologies to create a new multiproduct facility design. BioProcess Int. 5: 74–80. Wong, R. (2004). Disposable assemblies in biopharmaceutical production. BioProcess Int. 9 (Suppl.): 36–38. Bestwick, D. and Colton, R. (2009). Extractables and leachables from single‐use disposables. BioProcess Int. 1 (Suppl.): 88–94. DePalma A. (2017). Extractables and leachables: standardizing approaches to manage the risk. https:// bioprocessintl.com/analytical/leachables‐extractables‐ particulates/extractables‐and‐leachables‐standardizing‐ approaches‐to‐manage‐the‐risk (accessed 11 November 2018). Vogel, J. and Wormuth, K. (2017). Particulate contamination in single‐use systems: challenges of detection, measurement, and continuous

23

24

25

26

27

28

29

30

31

32

33

improvement. https://bioprocessintl.com/analytical/ leachables‐extractables‐particulates/particulate‐ contamination‐single‐use‐systems‐challenges‐ detection‐measurement‐continuous‐improvement (accessed 14 November 2018). Wormuth, K. (2018). Particulate contamination in single‐use systems: real versus perceived risk. ECI Conference “Single‐Use Technologies IIII: Scientific and Technological Advancements” (23–26 September). Snowbird, Utah. Bender, J. (2007). Protein concentration. In: Medicines from Animal Cell Culture (ed. G. Stacey and J. Davis), 331–346. Chichester West Sussex: Wiley. Wilson, M. (2007). Purification methods: protein concentration. In: Medicines from Animal Cell Culture (ed. G. Stacey and J. Davis), 347–370. Chichester, West Sussex: Wiley. Schmidt, F.R. (2007). From gene to product: the advantage of integrative biotechnology. In: Handbook of Pharmaceutical Biotechnology (ed. S.C. Gad), 1–52. Hoboken, NJ: Wiley. Shukla, A., Hubbard, B., Tressel, T. et al. (2007). Downstream processing of monoclonal antibodies – application of platform approaches. J. Chromatogr. B Analyt. Technol. Biomed Life Sci. 848 (1): 28–39. Jacobi, A., Eckermann, C., and Ambrosius, D. (2007). Developing an antibody purification process. In: Bioseparation and Bioprocessing – a Handbook, vol. 2 (ed. G. Subramanian), 431–457. Weinheim: Wiley‐VCH. Strube, J., Sommerfeld, S., and Lohrmann, M. (2007). Process development and optimization for biotechnology production – monoclonal antibodies. In: Bioseparation and Bioprocessing – a Handbook, vol. 1 (ed. G. Subramanian), 65–99. Weinheim: Wiley‐VCH. Phillips, M.W., Bolton, G., Krishnan, M. et al. (2007). Virus filtration process design and implementation. In: Process scale Bioseparations for the Biopharmaceutical Industry (ed. A.A. Shukla, M.R. Etzel and S. Gadam), 333–365. Boca Raton: CRC Press. Eibl, R. and Eibl, D. (2010). Antibody manufacture, disposable systems. In: Encyclopedia of Industrial Biotechnology: Bioprocess, Bioseparation and Cell Technology, vol. 1 (ed. M. Flickinger), 344–351. New York, Chichester, UK: Wiley. Paldus, B.A., Langer, E., Blomberg, M. et al. (2016). Single‐use/disposables technologies and equipment roundtable. https://www.americanpharmaceuticalreview. com/Featured‐Articles/184450‐Single‐Use‐Disposables‐ Technologies‐and‐Equipment‐Roundtable (accessed 18 November 2018). Pora, H. and Rawlings, B. (2009). A user’s checklist for introducing single‐use components into process systems. BioProcessInt. 4: 9–16.

­  References

34 Monge, M. (2009). Disposables: vendor selection &

35

36

37

38

39

qualification, supply chain & purchasing strategies: disposables in biopharmaceutical development and manufacturing. APV Conference “Lessons Learned in Disposable Biopharmaceutical Development and Manufacturing” (4–5 November). Dessau, Germany. Sinclair, A. and Monge, M. (2008). Disposable technologies implementation: understanding and managing risks. http://www.biopharminternational. com/disposable‐technologies‐implementation‐ understanding‐and‐managing‐risks (accessed 18 January 2010). Sinclair, A. (2009). The maturation of the biomanufacturing industry. BioProcess Int. 1 (Suppl.): 95–96. Sinclair, A. and Monge M. (2009). User viewpoints on disposables implementation: what end users think about single‐use systems. http://www.biopharminternational. com/user‐viewpoints‐disposables‐implementation (accessed 18 January 2010). Merseburger, T., Pahl, I., Müller, D., and Tanner, M. (2014). A risk analysis for production processes with disposable bioreactors. Adv. Biochem. Eng. Biotechnol. 138: 273–288. Merseburger, T., Pahl, I., Müller, D., Tanner, M. (2015). Recommendation for a risk analysis for production

40

41

42

43

44

45

processes with disposable bioreactors. https://dechema. de/dechema_media/Downloads/Positionspapiere/ SingleUse_RiskAnalysis_2015.pdf (accessed 16 November 2018). Anant, J. (2018). Risk Assessment for Single‐Use Pharmaceutical Manufacturing Systems. Kenilworth: Merck & Co. BioPlan Associates (2016). Thirteenth report and survey of biopharmaceutical manufacturing capacity and production: a study of biotherapeutic developers and contract manufacturing organizations. Bn 978‐1934‐106‐28‐0. Badertscher, B., Eibl, R., Eibl, D. (2016). Single‐Use Technologie von A‐Z. http://a‐z‐singleuse.org (accessed 16 November 2018). Anderlei, T., Eibl, D., Eibl, R., et al. (2018). Facility of the future. https://dechema.de/dechema_media/ Downloads/Positionspapiere/SingleUse_ FoF+2018+engl.pdf (accessed 16 November 2018). Eibl, R. and Eibl, D. (2017). Single‐Use‐Systeme in der biopharmazeutischen Produktion. Pharm. Ind. 79 (5): 719–724. Eibl, D. and Eibl, R. (2017). Flexible biomanufacturing for the production of biotherapeutics. Pharm. Bioprocess. 5 (1): 001–002.

11

13

2 Types of Single‐Use Bag Systems and Integrity Testing Methods Jens Rumsfeld1 and Regine Eibl2 1 

Sartorius Stedim Biotech GmbH, Göttingen, Germany School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology, Zurich University of Applied Sciences, Wädenswil, Switzerland

2 

2.1 ­Introduction Media, buffers, starting materials, intermediate and finished products may have to be sampled, processed, stored, mixed, frozen, thawed, transported in‐house, or shipped during the manufacture of biopharmaceuticals. Single‐use bag systems have become accepted alterna­ tives and even state‐of‐the‐art technologies to the sam­ pling, storage, and transportation systems fabricated from glass or stainless steel, which have been traditionally used in these processes. Their usage was originally based on the technology associated with sterile intravenous medical applications, which commonly employ single‐ use bags, consisting of one‐layer films made from ­polyvinyl chloride or ethylene vinyl acetate (EVA), for the storage of blood or infusion solutions. Various bioprocess bags have been used in the research, development, and manufacture of biopharmaceuticals over decades [1–3]. Figure 2.1 summarizes the typical fields of application of single‐use bag systems in biopharmaceutical production processes. A thorough assessment of the potential risks to the process, product, and/or patient must always be com­ pleted prior to the implementation of single‐use bags in these numerous applications (see also Chapter 8). Single‐use bags are made from plastic films [4] (Chapter  8), the composition of which strongly influ­ ences their construction, performance characteristics, and production capabilities. The customer can select the bag depending on the design, volume, available ports, tubing, in‐line filters, connectors, technical function, and package. The majority of bag manufacturers offer both standard and customer‐specific bag solutions. In this chapter the most important types of single‐use bag systems designed for the handling of fluids and

­ owders are introduced, with particular focus on the p characteristics and typical applications of tank liners, two‐dimensional (2D) and three‐dimensional (3D) bags. In addition, the bag‐handling and container systems that are, with few exceptions, required for ­ trouble‐free use of single‐use bags are discussed. Finally, the primary considerations associated with bag‐based freezing and thawing of cells, intermediate and finished products are briefly described, and the increasing importance of container closure integrity testing is presented.

2.2 ­Bags for Fluid and Powder Handling 2.2.1  Tank Liners Tank liners are simple and open single‐use bags used to line container and transportation systems. In some cases they are not gamma‐sterilized. As shown in Figure 2.2, tank lin­ ers are used in open, single‐use systems in which the bag is placed in a container to form a lining [5]. The sole function of the container is to provide physical support. The function of the tank liner is to contain the liner content and to create a barrier between the liner content and the container wall and/or environment. Tank liners are well suited for mixing tasks, such as media and buffer preparation or in the formulation of drug prod­ ucts, in which large quantities of solids must be added and dissolved. Commercially available overhead mixers can readily be integrated because these systems are open. Tank liners are offered without or with ­bottom drain line.

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

14

2  Types of Single‐Use Bag Systems and Integrity Testing Methods

Typical bag applications in biopharmaceutical manufacturing > Storage and product hold: Powder, media, buffers, Water for injection (WFI), reagents and bulk products > Mixing: Media and buffer preparation, formulation of drug products > Shipping and transportation: Delivery of media, buffers, WFI, reagents and bulk products to point-of-use > Cell cultivation and fermentation: Inoculum production, transfer of inoculum, fluid management for media, harvest and fermentation waste water > Sampling: Process samples, reference samples > Recovery and clarification: Fluid management for microfiltration > Purification and concentration: Fluid management for ultrafiltration, diafiltration, virus inactivation, and process chromatography > Freezing and thawing: Manufacturing of intermediates, inoculum or purified bulk products > Filling and finishing: Fluid management for final filtration and filling

Figure 2.1  Summary of single‐use bag applications in biopharmaceutical manufacturing.

Figure 2.2  Tank liner in a drum. Source: Reproduced with permission of Sartorius Stedim Biotech.

2.2.2  Two‐Dimensional Bags 2.2.2.1  Bags for Fluid Handling

2D bags are used when small volumes of liquid must be handled. Liquid volumes between 20 ml and 50 l are typi­ cal. Films used are mostly based on polyethylene (PE) or ethyl vinyl acetate (EVA) fluid contact layers. Larger bags with volumes up to 200 l are available, but are exceptions. They are utilized for special applications like fermenta­ tion (see also Chapter 4) as in the case of Cellbag 200 l and

Flexsafe RM 200 l. These bags are produced from multi­ layer films based on different resin materials, depending on product family and manufacturer. The result is a flat chamber which has ports either face‐welded (Figure 2.3b) or in most cases end‐welded (Figure 2.3a). The number and positioning of ports and assemblies (tubing, pinch clamps, connectors, filters, etc.) are ­manufacturer specific and depend on the bag type. Most bag manufacturers offer a number of variations within one standardized series of bag types. As mentioned in the introduction, customized solutions are available. However, customization increases costs because of the additional complexity, engineering, and cost of manufac­ turing. The use of customized bags can be cost‐effective, if the bags are used in large quantities or for very specific applications requiring special technologies as well as for high‐value products. 2D bags are utilized either in a reclining or hanging position. In addition to individual single‐use bags, multi­ ple 2D bags, for example, for use in manifolds for sam­ pling, dispensing, and product hold, are also available. 2.2.2.2  Bags for Powder Handling

Bags designed for the handling of solids are funnel‐ shaped, flexible, and transparent. They are equipped with large sanitary fittings or aseptic transfer systems (see Chapter  5) at the funnel outlet (see Figure  2.4).

2.3  Bag‐Handling and Container Systems

Figure 2.3  Examples of 2D bags: (a) end‐ported bag and (b) face‐ported bag. Source: Reproduced with permission of Sartorius Stedim Biotech.

(a)

(b)

liners) the outer bag container has a support function. Films used are mostly based on PE fluid contact layers. Depending on the customers’ requirements, these bags can accommodate liquid up to 5000 l in volume. The 3D design results from the welding of appropriate multilayer films (see also Chapter 8). The most common designs are cylindrical/conical‐shaped (Figure 2.5b) or cube‐shaped/ rectangular (Figure  2.5a), which are available with and without bottom drain outlet. In contrast to 2D bags, 3D bags offer more flexibility with regard to the positioning and number of ports. Top, bottom, and/or face‐ported 3D bags are available com­ mercially. Furthermore, there is a wide selection of port size and port complexity. Finally, it is worth mentioning that 3D bags form the basis of all larger single‐use stor­ age and transportation systems, nearly all single‐use mixing systems (except small‐volume, wave‐mixed sys­ tems, see Chapter 3), and all stirred single‐use bag biore­ actors (see Figure 2.5c and Chapter 4). Figure 2.4  Thermo Scientific Powdertainer II with clamp and wash down line. Source: Reproduced with permission of Thermo Fisher Scientific.

Applications of these powder bags (as they are also known) include closed transfer and storage, as well as transport of dry powder media, buffer salts, pharmaceu­ tically active substances and adjuvants. They are avail­ able from various manufacturers in different designs with capacities up to 100 l [5–8]. 2.2.3  Three‐Dimensional Bags 3D bags are available for more complex applications and larger liquid volumes in which (as in the case of tank

2.3 ­Bag‐Handling and Container Systems 2.3.1  Bag‐Handling Systems Due to the inherent characteristics of film materials and the welding and production technologies utilized during their manufacture, single‐use bag systems have a limited mechanical load capacity. The tensile strength of the films, coupled with the strength of the welded seams, largely determines the maximum mechanical load capac­ ity, and thus the size, shape, and application of the bag. In addition, the material characteristics are strongly tem­ perature dependent. For example, some plastic films may fail in a brittle manner at low temperature.

15

16

2  Types of Single‐Use Bag Systems and Integrity Testing Methods

(a)

(b)

(c)

Figure 2.5  Examples of 3D bags: (a) 3D Flexsafe bag for rectangular tank, (b) 3D Flexsafe bag for cylindrical/drum tank, and (c) 3D bag, Flexsafe STR, for fermentation. Source: Reproduced with permission of Sartorius Stedim Biotech.

(a)

(b)

(c)

Figure 2.6  Selected bag‐handling systems for liquids: (a) plastic Flexsafe/Flexboy trays, (b) stainless steel Flexsafe/Flexboy tray, and (c) Flexboy/Flexsafe rack. Source: Reproduced with permission of Sartorius Stedim Biotech.

Apart from small 2D bags used at ambient tempera­ tures for in‐house storage, all systems require an outer container. However, even small‐volume, 2D bags can be difficult to handle and therefore a suitable handling sys­ tem is also recommended for these bags to ensure safe storage and/or transportation. In many cases, simple, stackable trays or racks made from plastics or stainless steel are suitable for 2D bags; however, very complex systems can also be obtained for specific applications.

Simple systems such as stackable trays with carrying handles and inserts for tube guides are shown in Figure 2.6a and b [9, 10]. Optimum accessibility and flex­ ibility are ensured through the use of modular, expanda­ ble rack systems (see Figure 2.6c), which can also be used to fill and store bag manifold assemblies. Special protection must be provided for bags in which liquids are frozen or frozen liquids are stored and trans­ ported. The port/bag joints are of particular concern, but

2.3  Bag‐Handling and Container Systems

Figure 2.7  Bag‐handling system for frozen liquids. Source: Reproduced with permission of Sartorius Stedim Biotech.

there is also a risk of brittle failure at the tube/fitting transition points. Various manufacturers have therefore developed specific bags, made from special films and correspondingly secure handling systems, which can be used at low temperatures (see Figure  2.7). Some bag‐ handling systems even withstand freezing temperatures [11–13]. The most suitable, single‐use bag systems for freezing and thawing currently available are discussed in Section 2.4. 2.3.2  Container Systems for in‐House Applications Bioprocess containers for in‐house applications are used, if larger and more complex bags must be handled at the place of manufacture (Figure 2.8). They are made from (a)

(b)

stainless steel or plastic and are available for cylindrical/ conical‐shaped bags and cube/rectangular‐shaped bags up to 3000 l, but are not intended for shipping. Modern cube‐shaped bioprocess containers are stackable and occupy the minimum space necessary (Figure  2.8b). They can be transported with pallet‐lifting trucks, pallet jacks, forklifts, or dollies. In most cases, containers which exceed volumes of 500 l are operated in situ [9]. Bioprocess containers for mixing applications (see also Chapter 3) can be additionally equipped with load cells for weighing, jackets for temperature adjustment, and can be combined with control units for single‐use sen­ sors like pH and conductivity as required. The tempera­ ture of the bag content can be controlled through the heating or cooling of the heat‐transfer fluid circulated within a double jacket. 2.3.3  Container Systems for Liquid Shipping There is a demand for safe, stable, and closed container systems when sterile liquids are shipped in single‐use bags. The factors affecting the selection of suitable sys­ tems include the distance of travel, the means of trans­ portation, the temperature sensitivity, the volume, and the possibility of foam formation during formulation [14]. Such systems are referred to as liquid shipping con­ tainers. Container systems for liquid shipping can also be used for in‐house applications, insofar as they meet the demands of the application. It is important to note that some commercially available systems are only suited for usage in noncritical areas. In particular, in the case of larger volumes, the liquid must be held in position and is not allowed to oscillate. Vendors of container systems for liquid shipping, such as (c)

Figure 2.8  Bioprocess container for in‐house applications: (a) 3D Stainless steel Palletank for storage, (b) 3D plastic Palletank, and (c) 3D stainless steel Palletank for in‐process fluid handling. Source: Reproduced with permission of Sartorius Stedim Biotech.

17

18

2  Types of Single‐Use Bag Systems and Integrity Testing Methods

Figure 2.9  Flexel 3D Palletank for shipping. Source: Reproduced with permission of Sartorius Stedim Biotech.

Thermo Fisher Scientific and Sartorius Stedim, offer appropriate devices, known as the Smartainer II Shipper and the Flexel 3D Palltank for shipping, respectively. These devices ensure safe and secure shipping of sterile liquids up to 500 l in volume as an adjustable cover plate holds the liquid in place and thus suppresses any liquid wave action during the shipping process (see Figure 2.9). There are dif­ ferent shipping qualifications for shipping systems for 2D as well as 3D bags available on the market at this time. For the qualification level, it is very important to distinguish between shipping by truck or air on the one hand and the value of the product on the other hand. Currently, shipping systems are routinely used for the supply of media and buffer, but also for intermediate and bulk products. The most advanced qualification level currently available on the market is based on the ASTM D4169 standard [15].

2.4 ­Single‐Use Bag Systems for Freezing and Thawing Frozen, biopharmaceutical intermediate and finished products remain stable over the long term, thereby facili­ tating storage and transportation. Moreover, the poten­ tial for undesired reactions and contamination are minimized and transportation under controlled and low temperatures is therefore possible [16]. The freezing sys­ tems used in biopharmaceutical manufacturing depend on the nature and quantity of the frozen cargo. While small plastic tubes and bags find application at the milli­ liter scale, bags as well as carboys are used at the L scale (1–20 l), and large, transportable, and controlled freezing containers made from stainless steel are utilized in large‐ scale systems (50–500 l) [13, 17–21]. Single‐use bags have proven to be appropriate for the freezing of biologics. The composition of the bags,

e­nabling safe usage at low temperatures, as well as the  availability of suitable bag‐handling systems (see Section  2.3.1) and controllable freezing and thawing apparatus, facilitates high freezing and thawing rates without compromising operating safety [21]. A further benefit of bag systems is the incorporation of manifold filling systems (see Chapter 5). As a consequence, a sim­ ple portioning of the liquid can be achieved prior to freez­ ing, thereby facilitating further processing. This permits the subsequent thawing of only the required quantity of product at any particular time. Furthermore, new appli­ cation fields for bags are being developed in combination with freezing processes, enabling long‐term storage and decoupling of process steps. For example, the procedure for the production of the working cell bank‐derived seed inoculum can be simplified and shortened. By doing so, large‐volume cell banks can be filled, frozen, and stored in bags, for example  [22]. A potential disadvantage of bags that should not be underestimated, in addition to the recognized problem of leachables and extractables (see Chapters 1, 8, and 11), is undetected damage to the frozen bag. Bag integrity cannot typically be confirmed until the contents of the bag have been thawed [23]. Bags suitable for cryopreservation can be frozen and thawed in traditional freezing and thawing systems, but it is not possible to control the freezing and thawing pro­ cesses if simple devices are used. Controlled freezing and thawing requires the use of advanced freezing and thaw­ ing systems, such as the Celsius‐Paks system from Sartorius Stedim. This was the first controllable freezing and thawing apparatus for use with bags with working vol­ umes between 30 ml and 16.6 l. When used in combina­ tion with a control unit, freeze–thaw modules, transfer carts, storage modules, and shipping units, biopharma­ ceutical liquids can be filled into bags, frozen under con­ trolled conditions, stored and/or shipped, control thawed, and further processed (see Figure 2.10) [13, 24]. The com­ pany Single Use Support also offers systems for freezing and thawing. The technology is based on a shell in which single‐use bags of different suppliers can be installed. The shell is then put into a freezing system based on horizontal stainless steel plates. The technology provides a scalable, automated, and controlled freezing and thawing process. The smallest system allows freezing of up to 4 l (4 × 1 l) and the second available system for up to 240 l (24 × 10 l) (see Figure 2.11). A further system for freezing of up to 300 l is announced to be launched to the market soon.

2.5 ­Container Closure Integrity Testing Container closure integrity (CCI) is an increasingly important topic for single‐use systems and is currently considered one of the key challenges for single‐use

2.5  Container Closure Integrity Testing

(a)

(b)

(c)

(d)

Figure 2.10  Sartorius Stedim freeze–thaw system: (a) 1 and 2 l Celsius‐Pak, Celsius‐Pak Carrier, 8.3 and 16.6 L Celsius‐Pak, (b) FT 100 freeze–thaw module with CU 5000 thermal control unit and transfer carts, (c) Celsius Shippable Storage Module (SSM) and SSM trolley, and (d) Celsius shipper and Celsius SSM shipper. Source: Reproduced with permission of Sartorius Stedim Biotech.

(a)

(b)

Figure 2.11  Single Use Support RoSS shell (a) and RoSS.pFTU for up to 240 l volume (b). Source: Reproduced with permission of Single Use Support.

t­ echnology in biopharmaceutical industry. In the future, ­single‐use processing will require more and more robust solutions and validated CCI to improve patient safety and ensure regulatory compliance. To date, no clear reg­ ulatory guideline exists which precisely describes the requirement, science and test method for single‐use sys­ tems. The following sections aim to establish such an overview. Two new ASTM [25] E55 work items on integ­ rity were kicked off  –  one as a standard practice for integrity assurance and testing (supplier and end user) as well as one, which serves as a test method for microbial ingress testing on single‐use systems. Additionally, the PDA TR27 (1998) pharmaceutical package integrity is under revision to include single use [26]. The BPSA (2017) design, control, and monitoring of single‐use sys­ tems for integrity assurance [27] already covers single‐ use technology.

For suppliers, there are several important aspects to ensure integrity for single‐use systems. First of all, a quality by design approach and validation at the develop­ ment phase is needed to guarantee consistent robustness of single‐use systems after a complex production pro­ cess. This includes film extrusion, component manufac­ turing, assembly, and in most cases gamma irradiation (see also Chapters 8 and 18). Critical production param­ eters as well as design spaces have to be defined and monitored for each production step to reach lot‐to‐lot consistency. Besides quality and process controls, imple­ mentation of physical supplier integrity testing and point of use testing can bring additional safety for critical applications. To develop such integrity tests, it is neces­ sary to understand liquid leakage and bacteria ingress mechanisms. In other words, the correlation between detection limits and liquid leaks as well as microbial

19

2  Types of Single‐Use Bag Systems and Integrity Testing Methods

ingress under process conditions may be established. There is, however, still a lack of understanding of the defect size causing liquid leaks and microbial ingress into single‐use systems under real‐life conditions. To develop a testing strategy, a look at the existing statistics from production and quality documentation is recommended, as a history of long‐term bag production ranging in the millions can surely help in defining risks, leak patterns, potential test methods, as well as detection limits. The science of CCI in relation to plastic films and ­single‐use assemblies used in biopharmaceutical manu­ facturing has not been studied in detail; however, research relating to CCI of final product packaging pro­ vides a good basis. Thus, this can serve as a starting point for the investigation into defect size and can play a key role in the development and correlation of physi­ cal test detection limits to liquid leak and microbial results. Published data and our experiences show that there is a close relationship between liquid leaks and microbial ingress (see Figure  2.12). According to the studies by Keller [28], leak sizes for liquid flow are not significantly different from leak sizes for sterility loss. Additionally, critical leak sizes depend on parameters like liquid surface tension (fluid characteristics) and pressure (process parameters). Microbial ingress was  not found with micro‐tubes of 2 μm under any ­conditions, while studies by Gibney [29] showed that a liquid  filled micro‐leak should more readily allow for the ­passage of microorganisms. Furthermore, the maximum allowable leakage limit (MALL), a definition referenced in USP [30], should be taken into consideration. Table  2.1 shows a variety of studies that have determined the MALL to fall in the range of 2–5 μm when using capillary tubes at neg­ ative and positive pressure conditions, and various test methodologies. Morton et al. [33] describe, that there is no transmission of microbes, if there is no liquid leak.

This observation is important because it shows that unless liquid can be observed to be leaking, microbial contamination of the process solution cannot occur. Based on these findings, the following steps should be taken: determination of the MALL under process condi­ tions, establishment of the correlation between liquid leak/microbial ingress and physical integrity testing, and, finally, development and validation of the physical test methods with detection limits to confirm the absence of liquid leak/microbial ingress in single‐use systems. The findings of Keller [28] and Gibney [29] also sug­ gest that a liquid leak depends on pressure, liquid surface tension, and defect diameter. This needs to be taken into account for development of tests, as the following exam­ ple shows. The static pressure in a filled 500 l bag with a column water height of approximately 0.7 m corresponds to a head pressure of 70 mbar. In consistency with the Gibney [29] model, the liquid leak size would be expected to be around 20 μm. In experiments a leak at 20 μm was detected, but not at 15 μm. Based on these results, the MALL was defined for these process conditions as 10 μm after the introduction of a safety margin. When looking into shipping applications, additional aspects have to be considered. Pressure pulses during air shipment of bags showed no impact on the defect size, as there was no significant gas headspace present in the bags, thus the differential pressure can have no influ­ ence. However, what does have an impact on the defect size is the acceleration in liquid shipping. Real‐world air and truck transportation recorded acceleration levels of up to 20 g. By conducting a theoretical calculation using the Gibney [29] model, by increasing the acceleration from 1 to 20 g the corresponding MALL would be reduced from 10 to 2 μm, at a liquid height of 0.7 m (or 70 mbar) in the 500 l bag during shipment. This means that a test with 2 μm detection limit would make sense for a shipping application. Figure 2.12  Relationship between liquid leak and microbial ingress. Source: Reproduced with permission of Sartorius Stedim Biotech. DI, distilled.

Threshold leak diameter for DI water vs microbial ingress

100

Liquid leak Threshold diameter (μm)

20

80 Microbial ingress 60 40 20 – 0

50

100

150 MBAR

200

250

300

2.5  Container Closure Integrity Testing

Table 2.1  Summary of findings. Test/author

Burell [31]

Post [32]

Keller [28]

Gibney [29]

Microbial test

Immersion

Immersion

Aerosol

Aerosol

Tube length

3 cm

1.5 cm

0.7 cm

0.7 cm

MALL liquid leaks

Not available

Not available

2 μm

5 μm

MALL microbial ingress

5 μm

5 μm

2 μm

5 μm

Pressure – mbar

−250/+300

−250/+300

−210/+210

−350

Material

Glass

Glass

Nickel

Nickel

Source: From Keller [28], Burell [31], Gibney [29], and Post [32].

1 BC-LT

Bag making

Final assembly

1. Bag chamber leak test (BC-LT) 100% BC-LT

2 D BC-LT

Sterilization and shipment

2 SIT

3 PoU-LT

2. Final product supplier integrity test (SIT)

3 D BC-LT

Pressure decay

100% SIT Helium test

ASTM F2095(6) Detection limit: 40 – 90 μm for 2D 50 – 100 μm for 3D

Detection limit: 2 μm

3. Point of use leak test (PoU-LT)

Pressure decay ASTM F209 Detection limit: 10 μm on 2D bags 100 – 200 μm on 3D bags

Figure 2.13  Sartorius Stedim integrity testing approach for single‐use assemblies for critical applications such as drug product or drug substance processing and storage. Source: Reproduced with permission of Sartorius Stedim Biotech.

To summarize, literature and studies based on films with laser‐drilled holes, model solutions and different test conditions have shown that results obtained with film materials fall into existing models for sterile packaging using micro‐tubes. They also indicate that 2 μm is most probably the MALL under any relevant process condi­ tions for liquid leak as well as microbial ingress. On the other hand, a test strategy also has to take into account the practical requirements as well as the technical and economic limitations of test methodologies during real

production by suppliers and by end users (e.g. produc­ tion cost, time, space and investments). In Figure 2.13, an integrity testing approach for single‐ use assemblies for critical applications is exemplarily shown. It is a combination of three testing steps inte­ grated in the production process of the single‐use assem­ bly by the supplier as well as in the operation process at the biopharmaceutical manufacturer. After bag chamber manufacture and before assembly, 100% of bag chamber leaks are tested with a bag chamber

21

22

2  Types of Single‐Use Bag Systems and Integrity Testing Methods

leak test (BC‐LT) using a pressure decay test method with a detection limit of 40–90 μm for 2D bag and 50–100 μm for 3D bag chambers (volume and tact time dependent). This test is based on the ASTM F2095 ­pressure decay method. Bag chambers are put between restraining plates to reduce the stress on the bag and prevent the bag from overexpansion. This provides a small inflation volume and therefore allows higher test pressure, for example, 500 mbar for 2D bags and 300 mbar for 3D bags. Porous spacers between the bag chamber and the restraining plates help avoid masking effects of potential leaks. Combining small‐volume, high‐test pressure and spacers provides a reproducible, accurate and sensitive test. While bag chambers failing this test are discarded, production history shows that failure rates are very low. After the final assembly of the bag chambers with components like tubings and connectors (see also Chapter  5), a supplier integrity test (SIT) using the Helium test method is performed (see also Chapter 8). The main advantages of the Helium test are the achiev­ able detection rate, a short test time, and the possibility to test assembled finished products. By increasing the test sensitivity to a detection level of 2 μm, the failure rate of finished products released to packaging and steriliza­ tion was reduced to around 0 ppm. Finally, to target a 0 ppm failure rate at the end user, a point of use leak test (PoU‐LT) can be conducted after unpacking and before filling at the end‐user’s facility. This test method is also based on a pressure drop test, employing porous spacers to avoid masking effects. For 2D bags, a specific test system covering the restraining plates can be used, leading to shorter test times and a desired detection limit of, for example, 10 μm. For 3D bags, the test has to be performed on the unfolded bag in the final container, which has an impact on test time and reachable detection limit (currently 100–200 μm). To

keep sterility, 2D and 3D bags have to have dedicated designs including air filters to be able to inflate the bags with sterile air and eject the air following testing and during filling. The test sensitivities of 2 μm applying the mentioned Helium SIT and 10 μm applying the pressure decay PoU‐LT both correlate to the MALL, determined under the process conditions in which bags are used.

2.6 ­Summary and Conclusions The multitude of single‐use bag systems for liquid and powder handling available on the market is described in this chapter. The use of 2D and 3D bioprocess bags has become well established in most areas of biopharmaceu­ tical manufacturing including research, development, and production. They have been developed from the simple storage devices, typically used up to now, in to instrumented bioprocess units [34]. The importance of single‐use bags as alternative solutions in traditional sampling, storage, mixing, transport, freezing and thaw­ ing systems as well as bioprocess units is increasing with the growing acceptance of single‐use devices and, in par­ ticular, the adoption of platform technologies. Bioprocess bags provide greater flexibility and enable simpler execution of multiproduct facilities in the biop­ harmaceutical industry compared to more traditional systems [35, 36]. It is predicted that their acceptance will continue to grow in the future, driven largely by the availability of improved materials and manufacturing technologies and better understanding of the profes­ sional handling of single‐use bags and their accessories. The standardization of bags and containers is, however, highly desirable, in particular, to facilitate vendor replacement, but progress towards this goal has so far been limited (see also Chapter 12).

Nomenclature BC‐LT Bag chamber leak test CCI Container closure integrity CCIT Container closure integrity test DI Distilled EVA Ethyl vinyl acetate MALL Maximum allowable leakage limit

PE Polyethylene PoU‐LT Point of use leak test SIT Supplier integrity test SSM Shippable storage module 2D Two‐dimensional 3D Three‐dimensional

­References 1 Haughney, H. and Hutchinson, J. (2004). A disposable

option for bovine serum filtration and packaging. BioProcess Int. 4 (9 Suppl): 2–5.

2 Wong, R. (2004). Disposable assemblies in

biopharmaceutical production: design, implementation and troubleshooting. BioProcess Int. 4 (9 Suppl): 36–38.

­  References

3 Eibl, R. and Eibl, D. (2009). Application of disposable

4

5

6 7

8

9

10

11

12

13

14

15

16

bag bioreactors in tissue engineering and for the production of therapeutic proteins. Adv. Biochem. Engin. Biotechnol. 112: 183–207. Barbaroux, M. and Sette, A. (2006). Properties of materials used in single‐use flexible containers: requirements and analysis. http://biopharminternational. findpharma.com/biopharm/article/articleDetail.jsp?id= 423541&sk=&date=&pageID=7 (accessed 28 December 2009). ThermoFisher Scientific. (2018). Powdertainer II BioProcess Container (BPC). https://www. thermofisher.com/us/en/home/life‐science/ bioproduction/single‐use‐bioprocessing/flexible‐ containment/standard‐bioprocess‐containers/ powdertainer‐II‐bioprocess‐container.html (accessed 21 December 2018). ATMI. (2009). ATMI Products. Catalog. Sartorius Stedim Biotech. (2009). Products process: bags, containers & fluid management systems. http:// www.sartorius‐stedim.com/index.php?id=6495 (accessed 8 December 2009). Dover Corporation. (2009). DoverPac® SF. http://www. doverpac.com/products_doverpac_sf.cfm (accessed 28 December 2009). Bean, B., Matthews, T., Daniel, N. et al. (2008). Guided wave radar at genentech: a novel technique for non‐ invasive volume measurement in disposable bioprocess bags: GWR may be a cheaper, more practical alternative to traditional methods. http://www. pharmamanufacturing.com/articles/2008/185.html (accessed 8 December 2009). Wang, E. (2006). Cryopreservation, storage and transportation of biological process intermediates. BioProcess Int. Ind. Yearbook 2006: 78–79. Thermo Fisher Scientific. (2009). Thermo scientific nalgene bioprocess bag management system. http:// www.nalgenelabware.com/features/featureDetail. asp?featureID=70 (accessed 28 December 2009). UFP Technologies. (2009). BioShell™. http://www. bio‐shell.com/gallery.html (accessed 28 December 2009). Sartorius Stedim Biotech. (2009). Products process: freeze and thaw systems. http://www.sartorius‐stedim. com/index.php?id=9547. (accessed 28 December 2009). Lok, M. and Blumenblat, S. (2007). Critical design aspects of single‐use systems: some points to consider for successful implementation. BioProcess Int. 5 (5 Suppl.): 28–31. ASTM (2016). ASTM D4169. Prüfung der Gebrauchseigenschaften von Versandbehältern und ‐ systemen. Beuth. Singh, S.K. (2007). Storage considerations as part of the formulation development program for biologics. Am. Pharm. Rev. 10 (3): 26–33.

17 Goldstein, A., Loesch, J., Mazzarella, K. et al. (2009).

18

19

20

21

22

23

24

25

26 27

28

29

Freeze bulk bags: a case study in disposables implementation: Genentech’s evaluation of single‐use technologies for bulk freeze‐thaw, storage, and transportation. http://biopharminternational. findpharma.com/biopharm/Disposables+Articles/ Freeze‐Bulk‐Bags‐A‐Case‐Study‐in‐Disposables‐Imple/ ArticleStandard/Article/detail/637583 (accessed 28 December 2009). Zeta. (2009). FreezeContainer®. http://www.zeta.com/ DE/Produkte/Freeze‐Thaw‐Systeme (accessed 28 December 2009). Singh, S.K., Kolhe, P., Wang, W., and Nema, S. (2009). Large‐scale freezing of biologics: a practitioner’s review, part 1: fundamental aspects. BioProcess Int. 7 (9): 32–44. Singh, S.K., Kolhe, P., Wang, W., and Nema, S. (2009). Large‐scale freezing of biologics: a practitioner’s review, part 2: practical advice. BioProcess Int. 7 (10): 34–42. Rathore, N. and Rajan, R.S. (2008). Current perspectives on stability of protein drug products during formulation, fill and finish operations. Biotechnol. Prog. 24 (3): 504–514. Wolfgram, I., Lombriser, R., Neubauer, P. et al. (2017). Perfusion‐based production of inoculum cultures for large‐volume cell banks. Continuous Biomanufacturing (26–28 June). Oxford, UK. Lam, P. and Sane, S. (2007). Design and testing of a prototype large‐scale bag freeze‐thaw system: the development of a large‐scale bag freeze‐thaw system will have many benefits for the biopharmaceutical industry. http://biopharminternational.findpharma. com/biopharm/Disposables/Design‐and‐Testing‐of‐a‐ Prototype‐Large‐Scale‐Bag‐/ArticleStandard/Article/ detail/473322 (accessed 28 December 2009). Weidner, J. and Jimenez, F. (2008). Scale‐up case study for long term storage of a process intermediate in bags. http://www.americanpharmaceuticalreview.com/ ViewArticle.aspx?ContentID=3486 (accessed 13 January 2010). ASTM International. (2013). ASTM F2095 standard test methods for pressure decay leak test for flexible packages with and without restraining plates. Parenteral Drug Association. (1998). PDA TR27 pharmaceutical package integrity. Bio‐Process Systems Alliance. (2017). Design, control, and monitoring of SUS for integrity assurance. http:// bpsalliance.org (accessed 26 November 2018). Keller, S. (1998). Determination of the leak size critical to package sterility Maintenance. PhD Thesis. Virginia Polytechnic Institute and State University, VA. Gibney, M. (2000). Predicting package defects: quantification of critical leak size. Master Thesis. Virginia Polytechnic Institute and State University, VA.

23

24

2  Types of Single‐Use Bag Systems and Integrity Testing Methods

30 U.S. Pharmacopeia. (2015). USP Sterile

product – package integrity evaluation. 31 Burell, L., Carver, M., Demuth, G., and Lambert, W. (2000). Development of a dye ingress method to assess container‐ closure integrity: correlation to microbial ingress. PDA J. Pharm. Sci. Tech. 54 (6): 449–455. 2 Post, E. (2011). Container closure integrity test (CCIT): 3 cross‐validation of the microbiological vs a physico‐ chemical CCIT. PDA Europe Conference (15–18 March). Berlin, Germany. 3 Morton, D.K., Lordi, N.G., Troutman, L.H., and 3 Ambrosio, T.J. (1989). Quantitative and mechanistic

measurements of container/closure integrity: bubble, liquid, and microbial leakage tests. J. Parenter. Sci. Technol. 43 (4): 104–108. 34 DePalma, A. (2006). Bright sky for single‐use bioprocess products. GEN 26 (3): 50–57. 5 Langer, E.S. and Price, B.J. (2007). Biopharmaceutical 3 disposables as a disruptive future technology. BioPharm Int. 20 (6): 48–56. 36 Williamson, C., Fitzgerald, R., and Shukla, A.A. (2009). Strategies for implementing a BPC in commercial biologics manufacturing. BioProcess Int. 7 (10): 24–33.

25

3 Mixing Systems for Single‐Use Sören Werner1, Matthias Kraume2, and Dieter Eibl1 1

 School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology, Zurich University of Applied Sciences, Wädenswil, Switzerland  Technische Universität Berlin, Fakultät III Verfahrenstechnik, Lehrstuhl Verfahrenstechnik, Berlin, Germany

2

3.1 ­Introduction

3.2 ­The Mixing Process

Mixing constitutes one of the most critical process ­operations used in the manufacture of biopharmaceuti­ cals [1], and is utilized to achieve homogenization, ­suspension, dispersion (liquid–liquid and gas–liquid), and heat exchange. Such unit operations are often neces­ sary in medium and buffer preparation, cell cultivation procedures, during fermentations, in final formulation, and in the filling of biologicals. Mixing procedures may also be required in order to avoid sedimentation and demixing, or to achieve temperature shifts during harvest, concentration, purification, formulation, and ­ filling [2–4]. Conventional mixing devices fabricated from stainless steel currently represent the gold standard in biophar­ maceutical manufacturing [5]. The increasing interest in disposable technology within the biopharmaceutical industry, due to its numerous advantages [6–14] and the fewer constraints [15] compared to conventional steel, has resulted in the development of single‐use bag mixing systems up to a working volume of 5000 l. The customer can choose between numerous, single‐use bag mixers, which differ in their scale, mixing principle and their level of inherent cleanliness, instrumentation, and automation. Following a brief introduction to the main engineer­ ing aspects of mixing processes, the application of mixing in the manufacture of biopharmaceuticals is discussed. Single‐use bag mixing systems are catego­ rized and the design features, working principles, advantages, limitations, and potential applications of commercially available, single‐use bag mixers are described and discussed.

3.2.1  Definition and Description The process of mixing constitutes the distribution of solid or fluid elements in a volume in which these ­elements differ in at least one property. The properties are defined by concentration, aggregation, particle size and shape, drop size and shape, temperature, viscosity, color, density, etc. The aim of mixing is the realization of a specific mixing degree and the production of an intermediate or end product. Moreover, mixing is a prerequisite for some subsequent reactions and can facilitate both heat and mass transfer [16, 17]. Mixing can be divided into three categories: distributive mixing, dispersive mixing, and diffusive mixing. Distributive mixing is the adjustment of properties leading to spatial uniformity of all components. Dispersive mixing is defined as the disintegration of agglomerates or lumps to the desired ultimate grain size of the solid particulates or the domain size (drops) of the immiscible fluids [18]. In the latter, the existing stable agglomerates are broken up. In contrast, diffusive mixing (which is less relevant in industrial processes) is characterized by an equilibrium concentration resulting from molecular diffusion [16]. Laminar mixing, often encountered in fluids with high viscosities, originates from a longitudinal mixing where fluid motion is dominated by linear viscous forces. Fluid particles flow along parallel streamlines in a time‐ independent manner. In order to achieve homogeneity, additional radial mixing transverse to the stream lines is  necessary, which can be achieved through shearing, expanding, compressing, kneading, and utilization of backflow and spiral‐flow [16, 19]. Finally, turbulent

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

26

3  Mixing Systems for Single‐Use

­ ixing provides the greatest effectiveness. This is due to m spatial and temporal flow fluctuations requiring the fluid particles to reorient continuously along Lagrangian tra­ jectories [19], which leads to an effective turbulent mass transfer. Mixing processes are generally described by statistical parameters such as mixture quality, mixing time, and residence time distribution. Furthermore, the dimen­ sionless Reynolds number and the specific power input are important parameters in determining mixing effi­ ciency. All these parameters (which are briefly summa­ rized below) are also used for scaling‐up and scaling‐down of mixing processes. 3.2.2  Mixing Quality In general, mixing quality is typically defined as the deviation of a measured property related to the average value. A number of indices are used to quantify mixing, which differ depending on the application [20–22]. Six specific mixing quality conditions are generally recog­ nized: completely unmixed systems, ideally homogene­ ous mixing, demixing, homogeneous random mixing, real mixing, and texture mixing [21]. In a completely unmixed system, all components are locally separated

(Figure 3.1a). An ideally mixed system (also known as an ideally homogeneous system) is similar to a crystal lattice, in which every component x adjoins the same number of components y (Figure 3.1b). As illustrated in Figure 3.1c, demixing is characterized by spatial separa­ tion of components commencing adjacent to the walls of the system. A homogenous random mixture (Figure 3.1d) constitutes a condition in which there is identical probability that component x or y will be encountered after sufficient mixing time, which is to be expected as mixing is a random process. In practice, the actual mixing condition (Figure 3.1e) in technical mix­ ing systems lies between the conditions shown in Figure 3.1b and d. Texture mixing (Figure 3.1f ) has no importance in biomanufacturing. 3.2.3  Mixing Time Mixing time is the second parameter used to quantify mixing efficiency and represents the time necessary to reach a defined mixing quality. A degree of mixing of 95% is often assumed to represent an appropriate perfor­ mance in industrial mixing systems. But higher demands exist in special applications. The determination of the mixing time is described elsewhere [23].

(a)

(b)

(c)

(d)

(e)

(f)

Figure 3.1  Mixing conditions describing mixing quality. (a) Completely unmixed system, (b) ideally homogenous mixing, (c) demixing, (d) homogeneous random mixing, (e) real mixing, and (f ) texture mixing. Source: Adapted from Kraume [21]. Reproduced with permission of John Wiley & Sons.

3.3  Single‐Use Bag Mixing Systems

3.2.4  Residence Time Distribution Residence time distribution, introduced by Danckwerts [24], describes the probability of elements to remain within a continuously operated apparatus for a certain time. Thus, spatial and temporal mixing can be evalu­ ated [25]. 3.2.5  Reynolds Number The Reynolds number describing the ratio of inertial forces and viscous forces is named after the physicist Osborne Reynolds. It is used to analyze the fluid flow regime and is regarded as one of the most important parameters in model theory. For example, laminar flow occurs at low Reynolds numbers and is characterized by smooth and constant fluid flow. Turbulent flow develops at high Reynolds numbers. Unlike laminar flow in which viscous forces prevail, turbulent flow is characterized by inertial forces, which tend to produce flow instabilities, such as random eddies. 3.2.6  Specific Power Input The specific power input, defined as the mass‐ or ­volume‐related quantity of power introduced into a system by a drive mechanism, which can be mechanical, hydrau­ lic, or pneumatic, is a further, critical parameter used in optimizing the design of mixing systems, in determining their efficiency and also in scaling‐up [26]. A high spe­ cific power input will usually result in short mixing times. However, the shear stress in such a mixture will

typically be high and the temperature will tend to increase, both of which could be detrimental during the manufacture of both intermediate and final products, and, in particular, when utilizing single‐use bag mixing systems. The power input can be determined experimen­ tally, however, often special adaptation of the system is necessary [23].

3.3 ­Single‐Use Bag Mixing Systems 3.3.1  Overview and Classification Commercially available, single‐use bag mixing systems are categorized in Figure 3.2 based on the mechanism of power input and their working principle. In general, two main classes of single‐use mixing systems are currently available, namely mechanically and hydraulically driven. Here, the mechanically driven systems play the major role. As sug­ gested by the name, the power input required for mass and heat transfer in mechanically driven, single‐use mixers is provided mechanically. This class of mixers can be further subdivided into single‐use mixers with rotating stirrers, those with nonrotating stirrers producing a stirring effect through a tumbling motion, and those with oscillating devices such as vibrating disks. Single‐use mixers with rotating stirrers are either mechanically coupled to the shaft (with a simple seal) or magnetically coupled. In the case of tumbling and vibrating shafts, the shaft (or at least a sleeve) is directly welded to the single‐use mixing bag. Hydraulically driven, single‐use mixing devices are simpler in design than their mechanically driven

Single-use mixing devices

Mechanically driven

Rotating stirrer

Levitated

Magnetically coupled

Hydraulically driven

Tumbling stirrer

Mechanically coupled with sealing

Sealed shaft

Figure 3.2  Classification of single‐use bag mixing systems.

Oscillating

Sealed sleeve with wand

Vibrating disk or septum

Wave

Pump

27

28

3  Mixing Systems for Single‐Use

Table 3.1  Overview of currently existing bag mixing systems. Company

Mixer

Volume

Working principle

References

Pall

LevMixer System

30–1000 l

Stirred, levitated

[27]

Sartorius Stedim Biotech

Flexel LevMixer

50–1000 l

Stirred, levitated

[28]

Ecell

Polyethylene 3D Single Use Mixing System

50–1000 l

Stirred, magnetically coupled

[29]

GE Healthcare

Xcellerex XDM Quad

50–1000 l

Stirred, magnetically coupled

[30]

GE Healthcare

Xcellerex XDUO 2500

2500 l

Stirred, magnetically coupled

[30]

GE Healthcare

Xcellerex XDUO Quad

100–1000 l

Stirred, magnetically coupled

[30]

JM BioConnect

JetMixer Single‐Use Mixing System

10–1000 l

Stirred, magnetically coupled

[31]

Merck

Mobius Power MIX

100–3000 l

Stirred, magnetically coupled

[32]

Merck

Mobius MIX

10–1000 l

Stirred, magnetically coupled

[32]

Pall

Magnetic Mixer

6–2000 l

Stirred, magnetically coupled

[27]

Pall

Jet Mixer

10–1000 l

Stirred, magnetically coupled

[27]

Sartorius Stedim Biotech

Flexel Magnetic Mixer

50–2000 l

Stirred, magnetically coupled

[28]

ABEC

CSR General Mixer

50–5000 l

Stirred, directly coupled

[33]

Pall

Allegro 50 l High Performance Mixer for Single‐Use Operations

50 l

Stirred, directly coupled

[27]

Pall

Allegro Single‐Use Mixers

50–1000 l

Stirred, directly coupled

[27]

Thermo Fisher Scientific

HyPerforma Single‐Use Mixer (S.U.M.)

50–2000 l

Stirred, directly coupled

[34]

CerCell

CellMiscelatore Single‐Use‐Mixer

0.5–13 l

Stirred, magnetic stirrer table

[35]

Parker

SpinBag

2–20 l

Stirred, magnetic stirrer table

[36]

Romynox

Mixed4Sure

5–50 l

Stirred, magnetic stirrer table

[37]

Pall

PadMixer‐System

5–1000 l

Tumbling

[27]

DrM

FUNDAMIX SU

1–1000 l

Oscillating, vertically disk

[38]

Meissner Filtration

Saltus Single‐Use Mixing System

200 l

Oscillating, vertically disk

[39]

Parker

PoGo G2

200 l

Oscillating, vertically disk

[36]

Thermo Fisher Scientific

imPULSE Single‐Use Mixer (S.U.M.)

30–5000 l

Oscillating, vertically disk

[34]

GE Healthcare

Wave Mixer

Up to 35 l

Oscillating, vertically wave

[30]

Pall

Wand Mixer

1–200 l

Oscillating, wand in sleeve

[27]

JM BioConnect

QuattroMix Bags

Up to 1000 l

Hydraulically

[31]

Saint‐Gobain

Single‐Use Bioprocess Recirculating Mixing System

200–3000 l

Hydraulically

[40]

Information last accessed on company webpages on 15 November 2018. Please note that the completeness of this information cannot be guaranteed.

c­ ounterparts. A recirculation loop driven by an external pump provides a fluid flow through the bag. Special injectors can be used to intensify the mixing effect. Table 3.1 illustrates the multiplicity of existing, single‐use bag mixing systems. 3.3.2  Mixing Systems with Rotating Stirrer Magnetically coupled stirrers, which offer the advantage of high sterility and hence also security due to their closed design, are the most frequently used (Figure 3.3). However, wear debris generated at sliding surfaces can

contaminate the mixture. By levitating the stirrer, as in the case of the LevTech technology (interchangeable superconducting drive unit and proprietary levitating stirrer), particle generation is completely eliminated due to the stirrer’s frictionless motion [41]. However, the transmitted torque is limited by the magnetic force (Figure 3.3). 3.3.2.1  Levitated Mixers

The LevMixer system from Pall is referred to in the lit­ erature as a noninvasive, single‐use bag mixing system [41]. In essence, it consists of a bag incorporating a

3.3  Single‐Use Bag Mixing Systems

(a)

(b)

M

(c)

M

M

Figure 3.3  Stirred mixing systems. (a) Levitated stirrer. The superconducting drive unit induces levitation and rotation of the impeller through the plastic liner of the bag. (b) Magnetically coupled stirrer. Magnetic force is used to rotate the stirrer, which is integrated inside the bag and attached to a bearing. (c) Directly coupled stirrer. The drive and the stirrer are directly coupled through a shaft, which is integrated into the bag with a sealing.

bottom‐mounted, levitating impeller designed for powder–liquid and liquid–liquid mixing applications. Powder–liquid mixing tests have confirmed that the ability to resuspend sedimented particles is limited due to the upper speed constraint of the levitated impeller of 180 rpm. However, the system is intended for final formulation as well as for product homogenization and product suspension. All product‐contacting surfaces are completely disposable [41]. The system is available in sizes of 30–2000 l with varying tanks (round, medium‐ density polyethylene [MDPE] retaining tank, round and/ or cubical stainless steel retaining tank) [27]. The Flexel Bag for LevMixer from Sartorius combines the LevTech levitated impeller licensed by Pall and the Sartorius Flexel Bag. It comprises a stainless steel, cube‐ shaped container with a door for ease of bag mounting. In addition, it has windows to enable observation of the mixing process, a drive unit for levitating or rotating the stirrer, and a single‐use bag with a center‐mounted mag­ netic stirrer. The system is available in volumes of 50, 100, 200, 400, 650, and 1000 l [27]. 3.3.2.2  Magnetic Mixers

ABEC’s Custom Single Run (CSR) product line is cur­ rently one of the largest available single‐use mixing sys­ tems on the market. It is available up to a volume of 5000 l. Based on companies’ experiments, CSR‐General Mixing systems deliver process performance comparable to stainless steel systems. The system is intended to be used in upstream and downstream processes. It can be

fully integrated with filtration systems including pumps and sensors [33]. The Magnetic Mixer from Pall comprises a mixing bag incorporating a bottom‐mounted magnetically driven impeller capable of providing high‐torque mixing for all powder–liquid and liquid–liquid mixing applications [42, 43]. The powder–liquid mixing performance was evaluated in a 160 l tank with up to 37 kg of diatomaceous earth powder, which was homogenously mixed after less than two minutes. Even fully sedimented powder was completely mixed after five minutes without the impeller stalling [43]. The impeller rides on a low‐friction, inert, bearing assembly designed to ensure low attrition. A low particle concentration of less than 5 particles per milli­ liter with a particle size smaller than 10 μm was achieved during trials. The system allows the mixing of very high powder loads in large liquid volumes [42]. All surfaces which come into contact with the product are completely disposable [42, 43]. The system is available in sizes of 50 and 100 l (MDPE retaining tank, centered stirrer), 200, 350, and 500 l (MDPE retaining tank, two ports for ­stirrer), and 1000 and 2000 l (stainless steel carrier, off‐centered stirrer) [27]. Flexel Bag for Magnetic Mixer from Sartorius is avail­ able in sizes of 50–2000 l working volume. The system is intended for buffer and media preparation. These appli­ cations require a strong vortex in order to mix the pow­ der, which tends to float at the liquid surface [27, 44, 45]. GE Healthcare offers the Xcellerex XDM Quad Mixing System, which comprises an integrated magnetic stirrer

29

30

3  Mixing Systems for Single‐Use

with compact motor, a bottom‐mounted disposable ­stirrer, and an irradiated United States Pharmacopeia (USP) Class VI, single‐use low‐density polyethylene bag. The coupling between the motor and the disposable ­stirrer is magnetic. The square configuration offers enhanced mixing efficiency through a natural baffling effect, and compact storage capability. The bottom is slanted to ensure a low residual volume after discharge. The square configuration minimizes wrinkling of the bag and thus ensures reproducible processing. In‐line mea­ surement and sampling capability are achieved via a mixer window and bag probe ports, sampling port, and thermometer well, which is used for noninvasive tem­ perature measurement. The system provides flexibility with two standard bag types available that include vary­ ing numbers of tubing line, connections, and sampling and sensing capabilities to accommodate a wide variety of applications. The bags have an integrated 3‐in. Tri‐ clamp port for powder addition. A translucent lid is included to protect the bag and this is equipped with a securing feature for bag powder addition, a port, and a hopper. The 500 and 1000 l versions are equipped with removable doors for easy bag loading. The system is also available in  either polypropylene or jacketed stainless steel for heating and cooling applications. XDM Quad Mixing systems are available in five standard sizes: 50, 100, 200, 500, and 1000 l [30]. The Xcellerex XDUO 2500 Mixer from GE Healthcare exhibits a dual drive for two stirrers, which are mag­ netically coupled and can be controlled independently. Automated pH adjustment enables equilibration of cell‐culture media and buffer preparation, or can be used for automated virus inactivation. The system’s nominal volume is 2500 l, and thus, larger scale both upstream and downstream applications can be accom­ plished. The smaller Xcellerex XDUO Quad Intelligent Single‐Use Mixing System from GE Healthcare is avail­ able from 100 to 1000 l volume. It combines onboard automation capabilities, including in‐line sensing and process control, with good mixing. It supports upstream and downstream applications for automated mixing of buffer, media, product, and intermediates, as well as other process fluids [30]. The Mobius Power MIX family of single‐use mixers from Merck is based on the company’s magnetically cou­ pled NovAseptic mixing technology, which is also used with stainless steel tanks. The system is available in sizes of 100, 200, and 500 l (high‐density polyethylene and jacketed stainless steel) and 1000, 2000, 2500, and 3000 l (jacketed stainless only). For pH measurement, two options are available: a probe port for insertion of a reus­ able standard PG 13.5 threaded probe, or an installed, pre‐calibrated, and presterilized Hamilton OneFerm Single‐Use pH Sensor.

The company Ecell from Korea offers a mixing system based on bags in sizes of 50, 100, 200, 500, and 1000 l. It comprises a single‐use bag with an integrated impeller, which is magnetically coupled and a stainless steel tank with drive and optional load cells [33]. The key feature of Pall’s JetMixer system is a cube‐ shaped disposable bag with a magnetically driven ­turbine, mounted to the bottom of the bag, which works as a centrifugal pump. The three‐dimensional (3D) recir­ culation loop established during operation is claimed to eliminate dead zones and ensures fast and efficient ­mixing [29]. The particle generation is only 1% of the permitted particle level specified by the USP for water for injection and reaches only 0.2 particles per milliliter (10 μm) and 0.05 particles per milliliter (25 μm) after 24 hours of mixing at 1000 rpm [29]. This system is available in sizes of 50, 200, 500, and 1000 l working volume [27]. Similar bags are available from JM BioConnect as JetMixer Bags in sizes from 1 to 1000 l. The bag with the integrated mixing device, which is a magnetically cou­ pled turbine secured to the bottom, works like a centrif­ ugal pump by drawing liquid in vertically and expelling it horizontally [31]. Besides larger scale mixing system, small‐scale sys­ tems agitated by a magnetic stirrer table are available. The mixing system CellMiscelatore from CerCell can be equipped with different impellers in order to achieve good mixing. The rigid single‐use vessel can be ordered in sizes between 0.5 and 13 l [35]. Another system for usage with a magnetic stirrer table comprises a single‐ use stirrer system, which can be introduced to Nalgene bottles. The system Mixed4Sure from SaniSure results in mixing volumes up to 50 l [37]. Parker has created a bag mixing system for usage with a magnetic stirrer table called the SpinBag. The 3D bag with an integrated stirrer bar can be fixed in a stainless steel holding stand for the bags [36]. 3.3.2.3  Mixers with Sealing

The new Allegro Single Use Mixer from Pall with a directly coupled, bottom‐mounted stirrer offers high‐ power mixing for applications with high powder loads. Due to its directly coupled shaft, it can be considered very clean, which also results in its potential application for formulation processes. The system is available in sizes of 50, 200, 500, and 1000 l, and offers the possibility to attach sensor probes for online measurement (e.g. conductivity, pH, and temperature). The four‐pitched blade impeller can rotate either clockwise‐ or anticlock­ wise and, thus, an upward and downward flow is possible depending on the application. Furthermore, the small 50 l version of the Allegro Single Use Mixer system is equipped with a special stirrer, which allows operation with very small volumes down to 2 l only.

3.3  Single‐Use Bag Mixing Systems

The HyPerforma Single‐Use Mixer (S.U.M.) from Thermo Fisher Scientific comprises a stainless steel con­ tainer and a single‐use bag. It uses an impeller linked to an overhead drive and is coupled by a sealed bearing assembly which maintains the integrity of the system. The mixing stirrer is installed off‐center. This mixer is intended for powder/liquid and liquid/liquid mixing and has sterile single‐use contact surfaces. It is available in sizes of 50, 200, 500, 1000, and 2000 l working volume [27]. 3.3.3  Mixing Systems with Tumbling Stirrer Due to the flexibility of the bag, seals are not incorpo­ rated into mixing systems with a tumbling paddle designed for single use. The movement of the tumbling stirrer is introduced into the mixing vessel from the top through an axle completely built into the single‐use bag (Figure 3.4). The PadMixer from Pall has a top‐mounted mixing paddle that provides the stirring action and ensures effective and uniform mixing in demanding applications, such as the preparation of highly concentrated solid/liq­ uid mixtures or the mixing of highly viscous liquids. The direction of rotation of the paddle can be alternated to improve mixing and the paddle speed is limited to a maximum of 150 rpm. Mixing of diatomaceous earth slurry of up to 45 kg per 120 l required less than five min­ utes, while resuspension of sedimented powder was completed in less than one minute [45]. The Pad‐Drive is available in sizes of 25, 50, 200, 500, and 1000 l working volume in various configurations (paddle speed and ­paddle rotation angle) [27]. M

Figure 3.4  Mixing system with a tumbling stirrer. Due to the tumbling motion of the stirrer, a sleeve can be used to integrate the stirrer and move it with the help of a shaft, which is connected to the sleeve externally.

3.3.4  Mixing Systems with Oscillating Devices Instead of rotating or tumbling a stirrer to induce fluid flow, a rotational oscillating motion can be utilized (Figure  3.5). The Pall WandMixer consists of a mixing bag incorporating a top‐mounted impeller capable of providing efficient mixing for all powder–liquid and ­liquid–liquid mixing applications. The impeller comprises a rotational oscillating wand inside an inert polymer sleeve, and is designed to ensure low attrition and total containment while operating effectively in a wide variety of mixing applications. The mixing of 22 kg of diatoma­ ceous earth powder in 180 l of water was completed in 30 seconds at 250 rpm. Resuspension of the settled pow­ der was completed in a comparable time [46]. All sur­ faces which come into contact with products are completely disposable. The WandMixer is available in sizes of 1–20 l working volume as a benchtop unit, and 10–200 l working volume as floor‐based unit [27]. The imPULSE S.U.M. from Thermo Fisher Scientific currently represents the second mixing system on the market, which is available in sizes up to 5000 l. The key feature of the mixing system is the mixing disk, which is fabricated from rigid, engineering polymers. Multiple, evenly distributed slots penetrate the disk. The under­ side of the disk incorporates pie‐shaped flaps. These flaps open as the disk moves up from the bottom of the mixing bag on the drive’s upstroke, allowing fluid to flow through the disk’s slots. The flaps close on the down stroke, forcing the liquid towards the bottom of the hexag­onal vessel and subsequently up the walls of the vessel. The mixing disk, flaps, polymer mixing shaft, and the shaft rolling diaphragm seal, which attaches to the bag film, are all disposable. The single‐use mixer has a sealed plastic shaft that extends outside the mixing bag and is engaged with the mixer drive. The elastomeric seal, which moves with the drive mechanism thus mini­ mizing the stress on the single‐use assembly, is welded to the bag film and crimped to the hard plastic shaft. A vent port, a powder port, and a water/liquid port are located on the top of the mixing bag. A weight cell is included for control drive parameters weight/volume dependent. The system is intended for upstream (media preparation, buffer preparation, and harvest vessels) and downstream (pooling and liquid transfer, product suspension, mixing and storing multiple batches, buffer preparation, and viral inactivation) mixing applications [34]. Meissner Filtration Products offers a mixing system called SALTUS based on a vibrating disk with conical orifices. Due to the oscillating movement and the conical orifices, liquid jets develop at the tapered end of the holes. Thus, an axial fluid flow pattern is achieved. The fre­ quency and amplitude of the vibration can be adjusted to

31

32

3  Mixing Systems for Single‐Use

(a)

(b) M

M

Figure 3.5  Oscillating mixing systems. (a) Oscillating wand in sleeve. A wand is introduced to a sleeve incorporated into the bag and moved rotationally so that it oscillates. (b) Oscillating disk. Shaft with a disk with orifices with or without flaps is directly integrated into the bag and oscillated vertically. (c) Wave. A bag is connected to a rocking platform, which oscillates vertically.

(c)

M

provide either vigorous or gentle mixing. The single‐use bag is inserted into a stainless steel support c­ ontainer, with the drive installed on top of the container. The bag is preas­ sembled with one or two rigid vibrating disks and with tubes, filters, and a sampling port and temperature mea­ surement. Due to the frictionless, oscillating movement of the disk, it is assumed that there is no particle generation, and thus that the SALTUS is ultraclean. The bag is available in a size of 200 l working volume, with the minimum work­ ing volume being 10% of the maximum filling level. Due to the bottom‐mounted draining port, no fluid remains in the bag after the ­contents have been discharged [39]. The FUNDAMIX SU system from DrM combines the FUNDAMIX high‐performance mixing technology, which is used in many stainless steel applications, with single‐use technology. The technology is based upon a disk which oscillates up and down. The polyethylene mixing plate is mounted to a shaft which, in turn, is com­ pletely sealed to the bag. The system is available in sizes from 1 to 2000 l [38]. The PoGo G2 Mixing System from Parker’s Mitos Biotech Business Unit is available in 200 l, with other sizes available at customer’s request. This system exhibits reciprocal mixing disk technology [36]. In WAVE Mixers from GE Healthcare, mixing is achieved through horizontal oscillation of the bag

which is fixed in a rocker unit. The rocking motion is extremely efficient in generating waves, and the wave‐ induced motion in the bag causes large volumes of fluid to move, facilitating dispersion of solids. The optimum operating parameters depend on the combination of the container geometry, bag support, filling volume, rocking angle, rocking rate, and the characteristics of the mixture (solids, foam, etc.). Standard systems are available for 20 and 50 l bags, which can be used to mix volumes from 1 to 35 l of liquid. Larger systems of up to 500 l working volume are also available. Waves must be propagated in the bag to achieve efficient mixing and,  therefore, the bag should not be completely full. Typically, about 75% of the total bag volume can be uti­ lized. The bag is made of fatigue‐resistant materials and is mounted in such a way that creasing of the bag struc­ ture is minimized [30]. 3.3.5  Hydraulically Driven Mixing Systems In general, a hydraulically mixed system comprises a vessel or a bag and a circulation loop, which is pumped (Figure 3.6). Two systems are commercially available today. The Bioprocess Recirculating Mixing System from Saint‐Gobain represents a simple, single‐use mixing alter­ native for biopharmaceutical applications with working

­  References

3.4 ­Summary and Conclusions

Figure 3.6  Hydraulically driven mixing systems. An external pump moves the fluid.

volumes of 200, 500, 1000, 2000, and 3000 l. An externally mounted pump recirculates and mixes the fluid in a sin­ gle‐use bag system. The pump utilizes magnetic levitation technology which results in zero particle generation. Through a powder/liquid venturi port in the external loop, powder or liquid additions can be accomplished. The mixing system’s application functions as buffer and media preparation and storage, cell culture and centrifuge transport, harvest hold and clarification operations, chro­ matography and concentration operations, and product pooling and media distribution [40]. The QuattroMix Bags from JM BioConnect combine a single‐use bag with a single‐use membrane pump from QuattroFlow. The bags are available in varying sizes up to maximum volume of 1000 l. The speciality of the sys­ tem is the inlet: it consists of a specially designed plat with 16 nozzles, which leads to a smooth fluid stream inside the bag [31].

The working principles of current, single‐use bag mixing systems correlate well with those of their reusable coun­ terparts, although not all principles, designs, and sizes are transferable. This disparity arises from constraints due to the construction and manufacture of single‐use bags (see also Chapter 8). Single‐use mixers are limited in terms of their geometrical accuracy, stability, and the availability of specific shapes and sizes, due to the mate­ rials and m ­ anufacturing technologies used in their fabri­ cation. Not ­ surprisingly, most working principles of single‐use bag mixing systems are also to be found in single‐use bioreactors (see also Chapter 4). Tasks such as homogenization, suspension, gas–liquid dispersion and heat exchange ­(electric, double jacket, or external heat exchanger) are fundamental operations required both for successful mixing processes and for optimized culti­ vation procedures. Over the last few years, single‐use mixing systems have become increasingly important in biomanufacturing, and are now well accepted in all stages of the production process: upstream, downstream, in formulation, and fill­ ing. In addition to those mixing systems that incorporate stirrers, which represent the majority of single‐use bag mixing systems available, the customer can also select from mixers with oscillating bags, disks, or septa in addi­ tion to hydraulically driven systems. The choice of the most suitable type of single‐use bag mixing system is pri­ marily a function of the mixing tasks to be performed and the desired level of cleanliness. Although the handling and usage of single‐use bag mix­ ing systems is generally easier and more convenient than that of their stainless steel counterparts and their power input is lower, their mixing efficiency is also generally lower. Finally, single‐use solutions are not yet available for highly viscous media requiring high power input, for agi­ tators positioned near to the vessel wall (e.g. anchor stirrer and spiral stirrer), and for solid ­dispersion applications.

Nomenclature CSR Custom Single Run MDPE Medium‐density polyethylene

USP United States Pharmacopeia 3D Three‐dimensional

­References 1 DePalma, A. (2005). Liquid mixing: solid challenges.

http://www.pharmamanufacturing.com/ articles/2005/297.html (accessed 25 January 2010). 2 Agalloco, J. and Akers, J. (2008). Sterile product manufacture. In: Pharmaceutical Manufacturing

Handbook – Production and Processes (ed. S.C. Gad), 99–136. Hoboken: John Wiley & Sons, Inc. 3 Behme, S. (2009). Manufacturing of Pharmaceutical Proteins. Weinheim: Wiley‐VCH Verlag GmbH & Co. KGaA.

33

34

3  Mixing Systems for Single‐Use

4 Heinzle, E., Biwer, A.P., and Cooney, C.L. (2006).

5 6

7

8 9 10 11 12

13

14

15

16

17

18

19

Monoclonal antibodies. In: Development of Sustainable Bioprocesses – Modelling and Assessment (ed. E. Heinzle, A.P. Biwer and C.L. Cooney), 241–260. Chichester: John Wiley & Sons Ltd. Morrow, K. (2006). Disposable bioreactors gaining favor. GEN 26 (12): 42–45. Rader, R.A. and Langer, E. (2018). Single‐use bioprocessing equipment trends and adoption by CMOs. BioPharm Int. 31 (11): 10–13. Charles, I., Lee, J., and Dasarathy, Y. (2007). Single‐use technologies – a contract biomanufacturer’s perspective. BioPharm Int. 9 (Suppl): 31–36. D’Aquino, R. (2006). Bioprocessing systems go disposable. Chem. Eng. Prog. 102 (5): 8–11. DePalma, A. (2004). Bioprocessing on the way to total‐disposability manufacture. GEN 24 (3): 40–41. Glaser, V. (2005). Disposable bioreactors become standard fare. GEN 25 (14): 80–81. Kranjac, D. (2004). Validation of bioreactors: disposable vs. reusable. BioProcess Int. Industry Yearbook 86. Lewcook, A. (2007). Disposables crucial in future of patient‐specific meds. http://www.in‐ pharmatechnologist.com/Processing‐QC/Disposables‐ crucial‐in‐future‐of‐patient‐specific‐meds (accessed 25 January 2010). Sinclair, A. and Monge, M. (2009). Evaluating disposable mixing systems. Biopharm Int. 22 (2): 24–29. Singh, V. (2004). Bioprocessing tutorial: mixing in large disposable containers. Gen. Eng. Biotechnol. News. 24 (3): 42–43. DeGrazio FL. (2006). The importance of leachables and extractables testing for a successful product launch. http://pharmtech.findpharma.com/pharmtech/Article/ The‐Importance‐of‐Leachables‐and‐Extractables‐Test/ ArticleStandard/Article/detail/482447 (accessed 25 January 2010). Pahl, M. (2003). Mischtechnik, Aufgaben und Bedeutung. In: Mischen und Rühren – Grundlagen und moderne Verfahren (ed. M. Kraume), 1–19. Weinheim: Wiley‐VCH Verlag GmbH & Co. KGaA. Paul, E.L., Atiemo‐Obeng, V.A., and Kresta, S.M. (2004). Introduction. In: Handbook of Industrial Mixing – Science and Practice (ed. E.L. Paul, V.A. Atiemo‐Obeng and S.M. Kresta), xxxiii–lviii. Hoboken: John Wiley & Sons, Inc. Kresta, S.M. and Brodkey, R.S. (2004). Turbulence in mixing applications. In: Handbook of Industrial Mixing – Science and Practice (ed. E.L. Paul, V.A. Atiemo‐Obeng and S.M. Kresta), 19–88. Hoboken: John Wiley & Sons, Inc. Szalai, E.S., Alvarez, M.M., and Muzzio, F.J. (2004). Laminar mixing: a dynamic systems approach. In: Handbook of Industrial Mixing – Science and Practice

20 21

22 23

24

25

26

27

28

29

30

31

32

33 34

(ed. E.L. Paul, V.A. Atiemo‐Obeng and S.M. Kresta), 89–144. Hoboken: John Wiley & Sons, Inc. Boss, J. (1986). Evaluation of the homogeneity degree of a mixture. Bulk Solids Handling 6 (6): 1207–1215. Kraume, M. (2003). Mischen und Rühren – Grundlagen und moderne Verfahren. Weinheim: Wiley‐VCH Verlag GmbH & Co. KGaA. Zlokarnik, M. (2001). Stirring – Theory and Practice. Weinheim: Wiley‐VCH Verlag GmbH & Co. KGaA. Meusel, W., Löffelholz, C., Husemann, U. et al. (2016). Recommendations for process engineering characterisation of single‐use bioreactors and mixing systems by using experimental methods. https://dechema. de/dechema_media/Downloads/Positionspapiere/ SingleUse_ProcessEngineeringCaracterisation_2016. pdf (accessed 18 November 2018). Danckwerts, P.V. (1953). Continuous flow systems: distribution of residence times. Chem. Eng. Sci. 2: 1–13. Nauman, E.B. (2004). Residence time distributions. In: Handbook of Industrial Mixing – Science and Practice (ed. E.L. Paul, V.A. Atiemo‐Obeng and S.M. Kresta), 1–18. Hoboken: John Wiley & Sons, Inc. Zlokarnik, M. (2006). Scale‐up in Chemical Engineering. Weinheim: Wiley‐VCH Verlag GmbH & Co. KGaA. Pall. (2018). Webpage. https://shop.pall.com/us/en/ biotech/single‐use‐solutions/mixers (accessed 22 February 2019). Sartorius Stedim Biotech. (2018). Webpage. https:// www.sartorius.com/en/products/fluid‐management/ mixing (accessed 22 February 2019). Ecell. (2018). Webpage.http://www.ecellprocess.com/ bbs/content.php?co_id=sub0304 (accessed 22 February 2019). GE Healthcare. (2018). Webpage. https://www. gelifesciences.com/en/us/shop/liquid‐preparation‐and‐ management/single‐use‐mixers/mixing‐systems (accessed 22 February 2019). JM BioConnect. (2018). Webpage. http://www. jmbioconnect.com/jetmixing/ (accessed 22 February 2019). Merck. (2018). Webpage. http://www.merckmillipore. com/CH/de/products/biopharmaceutical‐ manufacturing/upstream‐processing/single‐use‐ manufacturing/mobius‐mixing‐systems/GBGb.qB. RM4AAAFDBEdUTxI9,nav (accessed 22 February 2019). ABEC. (2018). Webpage. http://abec.com/single‐use/ (accessed 22 February 2019). Thermo Fischer Scientific. (2018). Webpage. https:// www.thermofisher.com/ch/en/home/life‐science/ bioproduction/single‐use‐bioprocessing/single‐use‐ equipment/single‐use‐mixers.html (accessed 22 February 2019).

­  References

35 CerCell. (2018). Webpage. https://cercell.com/

36

37

38 39

40

41

42

products/configurable‐product/cellmiscelatore‐sum/ (accessed 22 February 2019). Parker. (2018). Webpage. https://www.parker.com/ literature/Domnick%20Hunter%20Process%20 Filtration%20Division/PAFD_literature/Life%20 Sciences%20Literature/Life%20Sciences%20 Product%20Datasheets/DS_P_SpinBag.pdf (accessed 22 February 2019). Romynox. (2018). Webpage. https://sanisure.com/ product/mixed4sure‐carboy‐mixing‐system/ (accessed 22 February 2019). DrM. (2018). Webpage. https://drm.ch/de/portfolio‐ item/single‐use‐mischer/ (accessed 22 February 2019). Meissner Filtration. (2018). Webpage https://www. meissner.com/de/produkte/saltus‐single‐use‐mixing‐ system (accessed 22 February 2019). Saint‐Gobain. (2018). Webpage. https://www. biopharm.saint‐gobain.com/single‐use‐ bioprocess‐recirculating‐mixing‐system (accessed 22 February 2019). Pall. (2015). Mixing of high powder loads using a LevMixer System. Application note. https://shop.pall. com/INTERSHOP/web/WFS/PALL‐PALLUS‐Site/en_US/‐ /USD/ViewProductAttachment‐OpenFile?UnitName= PALL&=;LocaleId=en_US&DirectoryPath=pdfs/ Biopharmaceuticals&FileName=USD3040‐LevMixer‐ Mixing‐High‐Powder‐Loads‐Application‐Note‐ GN15‐6189.pdf (accessed 22 February 2019). Pall. (2015). Particle generation in the Pall Magnetic Mixer. Application note. https://shop.pall.com/ INTERSHOP/web/WFS/PALL‐PALLUS‐Site/en_US/‐/ USD/ViewProductAttachment‐OpenFile?UnitName= PALL&=;LocaleId=en_US&DirectoryPath=pdfs/Bio pharmaceuticals&FileName=USD3030‐Magnetic‐

43

44

45

46

Mixer‐Particle‐Generation‐Application‐Note‐ GN15‐6147.pdf (accessed 22 February 2019). Pall. (2015). Mixing and resuspension of high powder loads using a Pall Magnetic Mixer. Application note. https://shop.pall.com/INTERSHOP/web/WFS/PALL‐ PALLUS‐Site/en_US/‐/USD/ViewProductAttachment‐ OpenFile?UnitName=PALL&=;LocaleId=en_US& DirectoryPath=pdfs/Biopharmaceuticals&FileName= USD3029‐Magnetic‐Mixer‐Mixing‐and‐Resuspension‐ High‐Powder‐Loads‐Application‐Note‐GN15‐6145.pdf (accessed 22 February 2019). Sartorius Stedim Biotech. (2018). 1,500 L and 2,500 L buffer and media preparations with Flexel for Magnetic Mixer. Application note. http://microsite.sartorius. com/fileadmin/sartorius_pdf/alle/biotech/Data_ Flexel_for_Magnatic_Mixer_50L‐2000L_SPT2023‐e. pdf (accessed 22 February 2019). Pall. (2015). Mixing a diatomaceous earth slurry using a PadMixer system. Application note. https://shop.pall. com/INTERSHOP/web/WFS/PALL‐PALLUS‐Site/ en_US/‐/USD/ViewProductAttachment‐OpenFile? UnitName=PALL&=;LocaleId=en_ US&DirectoryPath=pdfs/Biopharmaceuticals& FileName=USD3066‐PadMixer‐Mixing‐a‐Diatomaceous‐ Earth‐Slurry‐Application‐Note‐GN15‐6249.pdf (accessed 22 February 2019). Pall. (2015). Mixing a diatomaceous earth slurry using the Pall WandMixer. Application note. https://shop. pall.com/INTERSHOP/web/WFS/PALL‐PALLUS‐ Site/en_US/‐/USD/ViewProductAttachment‐OpenFile ?UnitName=PALL&=;LocaleId=en_ US&DirectoryPath=pdfs/Biopharmaceuticals&File Name=USD3052‐Wand‐Mixer‐Mixing‐Diatomaceous‐ Earth‐Application‐Note‐GN15‐6217.pdf (accessed 22 February 2019).

35

37

4 Single‐Use Bioreactors – An Overview Valentin Jossen, Regine Eibl, and Dieter Eibl School of Life Sciences and Facility Management, Institute of Chemistry and Biotechnology, Zurich University of Applied Sciences, Wädenswil, Switzerland

4.1 ­Introduction Stirred bioreactors manufactured from glass or stainless steel have been the systems of choice for the development and manufacture of biopharmaceuticals to date, irrespective of the organism being cultivated. Since their introduction, more than 60 years of bioreactor development have ensured that the user can now select between a multitude of different impeller types and aeration devices, and also has access to relevant engineering data such as fluid flow, mixing time, oxygen transfer efficiency rate, and residence time distribution [1]. This permits the identification of potential limitations of certain bioreactor designs, enables the improvement of process efficiency, facilitates comparison with other bioreactor types, and allows rapid scaling‐up and scaling‐down. Driven by strong international competition in the pharmaceutical industry and growing financial pressure in the health‐care sector, there has been a tendency over the last 15 years to move to single‐use bioreactors (SUBs) in modern biotechnological production processes [2]. Initially used in research to reduce development time when scouting for cell clones, culture medium, and process conditions, SUBs are now regarded as well‐ established in preclinical as well as clinical sample production and in the manufacture of small‐ to medium‐ volume products. The products typically represent high‐ value substances such as therapeutic monoclonal antibodies (mAbs), protein‐derived vaccines, therapeutic cells, and viruses, based on mammalian or insect cells [3–9]. In addition, the suitability of SUBs for production of plant cell culture‐based secondary metabolites as well as  recombinant proteins [10–12] and microbial products  [13–15] has been successfully demonstrated. For plant cell cultures, microbial cultures, and stem cells,  even specially designed SUB versions have been

developed [16–21]. Nevertheless, their current percentage application represents only a small fraction compared to those with animal cells. As outlined in Figure 4.1, the cultivation container in SUBs is intended for single‐use only and therefore cannot be reused. There are differences in the design of the cultivation container (its form, instrumentation, and scale) and the underlying power input in the SUB types (Figure 4.2). A categorization into static and dynamic SUB systems according to the type of power input has been recommended by different authors [3, 11, 12]. The predominant dynamic systems are characterized by improved mass and energy transfer and thus higher achievable cell densities and product titers. They are further subdivided into hydraulically driven, pneumatically driven, and mechanically driven bioreactors and their combinations, so‐called hybrid systems. Normally, the plastic cultivation containers are rigid display plates, tubes, cartridges, flasks, or cylindrical vessels at small volume (working volume at μl and ml range) and benchtop scale. Single‐use pilot and production‐scale bioreactors (two‐digit and three‐digit liter range) operate exclusively with inflatable flexible bags being cylindrically or cube‐shaped. The shape and the fixation of the bag are determined by a tray or a customized support container, which also comprises a tempering system in the form of a heater mat or double jacket. It is worth mentioning that not all sizes of cultivation containers are available for every SUB type and that they are not standardized and (thus) exchangeable for one and the same SUB type and scale. Furthermore, the container instrumentation differs in dependence on the scale and supplier of the SUB. As is obvious from Figure  4.2, there exist non‐, less‐, and highly

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

38

4  Single‐Use Bioreactors – An Overview

instrumented single‐use cultivation containers, whereby the container instrumentation is predetermined by the supplier. Both, single‐use and reusable sensors are installed in single‐use cultivation containers with their own measurement and control system. Because single‐use sensors are already implemented and sterilized by the supplier of the cultivation container, they are regarded as safer than their reusable counterparts. For this reason, many users prefer single‐use sensors, the current development stage of which is described in Chapter 6 of this textbook. The present chapter aims to give an overview of instrumented SUB types that are currently available commercially. Based on the milestones in the development of SUBs, we focus on the three most often applied types: wave‐mixed, stirred, and orbitally shaken versions. Their typical working principles are explained, possible fields of application are deduced, and trends in their selection are summarized.

> Meets USP Class VI and ISO 10993 specifications > Is sterilized by gamma radiation, customized, and validated > Is non-reusable

Figure 4.1  Common features of single‐use bioreactors.

Differences

Scale

Power input

The development of SUBs can be described in three phases: (i) the early beginnings, (ii) the establishment of disposable membrane bioreactors, multitray cell culture systems, and the first bag bioreactors, and (iii) the expansion of wave‐mixed, stirred, orbitally shaken, and further SUBs. As depicted in Figure  4.3, early cultivations in ­disposable devices date back nearly 60 years. At this time, glass Petri dishes had already been replaced by their plastic counterparts in a few microbial laboratories. In 1963, Falch and Heden [22] at the Karolinska Institute in Stockholm reported the successful application of shaken tetrahedron bags made of polypropylene and Teflon, which had been made for the first time in their own laboratory. They observed excellent growth of Bacillus subtilis, Escherichia coli, and Serratia marcescens cells (50 ml working volume).

In 1972, hollow fiber technology was introduced by Knazek and his team [23]. This permitted the development of numerous hollow fiber bioreactor systems (HFBSs) for animal cells grown anchorage‐dependent or in suspensions, such as the Cellmax HFBS (FiberCell Systems), the AcuSyst‐HFBSs (C3), the XCell HFBS (BioVest), and the Quantum Cell Expansion System (Terumo BCT). In these HFBSs, where power input is provided by double‐phase pumps, cells grow inside cartridges around a semipermeable hollow fiber membrane,

> Is made of FDA-approved polymeric materials

Instrumentation

4.2.1  Phase 1: Early Beginnings

4.2.2  Period 2: Establishment of Disposable Membrane Bioreactors, Multitray Cell Culture Systems, and the First Bag Bioreactors

The cultivation container:

Form

4.2 ­SUB History

Categorization into: > Rigid systems: plate, tube, cartridge, flask, vessel > Flexible systems: bag > Non-instrumented systems > Less-instrumented systems > Highly instrumented systems > Small-volume systems: screening scale > Medium-volume systems: benchtop and pilot scale > Large-volume systems: production scale > Static systems > Dynamic systems > Hydraulically driven > Pneumatically driven > Mechanically driven: wave-mixed (oscillated), stirred, orbitally shaken > Hybrid

Figure 4.2  Differences and categorization of single‐use bioreactors.

4.2  SUB History

Phase 1

Phase 2

Phase 3

From 1960s until early 1970s

From 1970s until early 1990s

> Single-use cultivation devices gain a foothold > Petri dishes > T-flasks > Roller bottles

> Development and use of disposable membrance bioreactors > Hollow fiber systems > Two-compartment systems

> First cultivation in shaken plastic bags

From late 1990s until today

> Static bags become fiber in cell expansion

> Wave-mixed, stirred, and orbitally shaken bioreactors expand in: > Screening experiments > Pre- and clinical sample productions > Seed inoculum productions > Small- and midvolume-scale manufacturing

> First cultivations in pneumatically driven and rocking bags

> Fully fiber automated microbioreactors available

> Multitray cell culture systems

Figure 4.3  Phases and milestones in single‐use bioreactor development.

with a molecular weight cutoff that is considerably less than the molecular weight of the target product. From perfusion around the membrane, cells are fed and the waste is removed. Oxygen enrichment is accomplished by a separate oxygenator module of fibers or silicon tubes. Hydraulically driven HFBSs were favored for in vitro productions of hybridoma‐derived mAbs in low‐ volume scale (100 mg to several grams) in the 1980s and  1990s [24–26]. Despite the high cell‐density levels achievable (107–109  cells/ml) and the high concentrations of high molecular‐weight secreted proteins, HFBSs have limitations with respect to the homogeneity of the culture environment and to scale‐up. The volume of the cultivation module is limited to 110 ml and scale‐up can only be achieved with the use of multiple modules or with an increase in the number of bioreactor systems [27]. For these reasons, HFBSs play a less significant role among SUBs today, although they are still popular in R&D (e.g. for expansion of human T cells and mesenchymal stem cells) [28–30] and in the manufacturing of in  vivo diagnostics [31, 32] that are required in small amounts of between 1 and 2 kg per year. In 1975, the company Nunc, working in cooperation with Bioferon (known today as Rentschler Biopharma), began to produce the Cell Factory from polystyrene [33]. This non‐instrumented, flask‐like culture system contains several trays stacked in parallel, one above the other in a single unit. Scale‐up could easily be achieved by increasing the number of trays up to a maximum of 40 [34]. Because this was a format suitable for industrial production, Cell Factories were used with a wide range of adherent animal cell lines [35–37]. They replaced

plastic roller bottles in the 1990s and have also proven to be suitable for the commercial Good Manufacturing Practice (GMP) manufacture of several vaccines (e.g. polio, rabies, rotavirus, and hepatitis‐A vaccines), therapeutic proteins (erythropoietin and interferon), human growth hormones, and human mesenchymal stem cells (hMSCs) [32, 33, 38–41]. In 1995, Osmotek introduced the LifeReactor. In this pneumatically driven bubble column bioreactor, mass and heat transfer are achieved by direct sparging of a conical‐shaped disposable culture bag (1.5–5 l working volume). The LifeReactor and its temporary immersion version, the Ebb‐and‐Flow Bioreactor, enabled advantageous growth of organ cultures of plant origin (meristematic clusters and somatic embryos) [42–45]. At the end of the 1990s, Curtis and his team recognized the suitability of inexpensive pneumatically driven plastic bags to propagate plant cells expressing secondary metabolites [46]. They designed a further plastic column with minimal instrumentation, the Plastic‐lined Bioreactor. The results of tests with suspension cells of Hyoscyamus muticus encouraged the group to scale‐up to 100 l working volume [47]. At the same time, two‐compartment dialysis membrane bioreactors were introduced into cell culture labs. As in the case with HFBSs, a semipermeable membrane separates the cells from the bulk of the medium and again permits diffusion of nutrients into the cell compartment with simultaneous removal of waste products. Both types, the MiniPerm (Greiner Bio One), which can be viewed as a modified roller bottle, and the T‐flask‐based CeLLine (INTEGRA Biosciences), must be kept in a CO2

39

40

4  Single‐Use Bioreactors – An Overview

incubator because they lack the degree of parameter control available in the previously described HFBSs. Efficient gas transfer for culture volumes ranging from 5 to 35 ml is ensured by a silicone membrane on the outside of both bioreactors. CeLLine and MiniPerm again enable high cell densities (>1 × 107  cells/ml) and high glycoprotein levels when growing animal and human cells [47, 48]. Both permit mAb production in the range of 100–500 mg, and can be operated for a period of several months [49, 50]. The popularity of the CeLLine for long‐term expansions of different cell types and screening experiments at the laboratory scale is well‐documented [51–57]. 4.2.3  Period 3: Expansion of Wave‐Mixed, Stirred, Orbitally Shaken, and Further SUB Types In 1998, the launch of the Wave Bioreactor 20, the first commercially available wave‐mixed bag bioreactor system at the laboratory scale, paved the way for the development of scalable, instrumented SUBs. Interestingly, the suitability of wave‐induced mixing for the cultivation of microorganisms and cell cultures in bags had already been described by Kybal, Vlcek, and Sikyta in two papers in 1976 and 1985 [58, 59]. The success of the Wave Bioreactor, originally designed to replace animal cell‐ based seed inoculum production [60], resulted in its scale‐up to 500 l working volume and the development of further wave‐mixed bag bioreactor types. In addition to the today’s commercially available wave‐mixed bioreactors listed in Table  4.1 and described in Section  4.3.1, customized versions were also designed by appliers such as the Wave and Undertow Bioreactor by specialists of Nestlé [16, 17, 61]. Since 2006, advocates of stirring have also had access to stirred bag bioreactors. Hyclone’s (today known as Thermo Fisher’s Scientific) S.U.B and the XDR‐ Disposable Stirred Tank Bioreactor from Xcellerex were the only systems available initially. In summer 2009, the Mobius CellReady 3 l bioreactor was introduced, which was the first stirred SUB at the benchtop scale [62, 63]. Now, it is possible to choose between a multitude of stirred SUBs. The reactors are mostly bubble aerated and operate with centrically mounted, rotating impellers in the majority. Different suppliers offer scalable platforms for stirred SUBs. The growing importance of orbitally shaken SUBs in the last two decades can be attributed to the increasing usage of shaken multiwell plates, as well as Erlenmeyers, and the favorable practical trials using the TubeSpin® technology [64–66]. Orbitally shaken SUBs have become the third largest group among available SUBs. This is ascribed to the discovery that orbital shaking is better tolerated by the production cells than originally assumed [67] and resulted finally in AmProtein’s CURRENT

Bioreactor line [68, 69] and Kühner’s OrbShake series [70, 71]. For more detailed information about orbitally shaken SUBs, the interested reader is referred to Section 4.3.3. In the early 2000s, the successful application of further mechanically driven SUBs, namely of the Saltus Vibromix Bioreactor [72, 73] and the Bayshake Bioreactor, [74, 75] was reported. In addition, pneumatically driven SUBs such as the Slug Bubble Bioreactor [16, 17, 76, 77], the Pneumatic Bioreactor System platform [78, 79], the CellMaker Regular [80], and the hybrid CellMaker Plus combining pneumatic and mechanic drive [81] have proven themselves in cultivation studies. As in the case with hydraulically driven, single‐use fixed‐bed systems, they have less importance than wave‐mixed, stirred, and orbitally shaken SUBs up until this point. Representatives of single‐use fixed‐bed, also referred to as packed‐bed, bioreactors such as the Eppendorf BioBLU Packed‐Bed Systems [9, 82, 83] or the iCELLis line from Pall [84–86] have been demonstrated to support the growth of shear sensitive cells (e.g. hMSCs) and adherent cells (e.g. Vero cells), which are still popular in many vaccine production processes. We would like to point out that the majority of cultivations executed with SUBs are those with suspension cells. Only HFBSs and packed‐bed bioreactors are dynamic bioreactors suitable for adherent production organisms. All other previously mentioned dynamic SUBs, and also the versions summarized in Table 4.1, must be operated with microcarriers providing the attachment matrix for adherent cells.

4.3 ­Comparison of the Current, Most Common SUB Types 4.3.1  Wave‐Mixed SUBs All wave‐mixed bioreactors shown in Table 4.1 operate with a bag consisting of a multilayer film. The bag is often made from polyethylene. Mixing within the bag is achieved through a wave‐induced mixing process resulting from the oscillating movement of the rocking platform. The wave characteristics inside the bag depend strongly on the bag shape/geometry, the rocking angle, the rocking rate, the bag translation, the filling volume (up to a maximum of 50%), and the fluid properties (liquid density and viscosity). The latter is more critical in plant cell cultivations where the fluid properties may change over the culture time [87]. Due to the directed wave movement, oxygen is introduced into the fluid from the headspace, resulting in a bubble‐free surface aeration which is beneficial for shear sensitive cells. This statement is also supported by the investigations of Werner et al. [88], who found, based on computational

Table 4.1  Overview of often used commercial single‐use bioreactors being instrumented and operating from ml to m3 scale.

Bioreactor type

Bioreactor brand

Vendor

Working volume (l)

Stirred with rigid container

CellVessel

CerCell

0.1–27

BactoVessel BioBLU®

0.1–27 Eppendorf

BioBLU®

0.067–3.75

Vessel with integrated optical DO (VisiWell), classical pH (FermProbe), and/or Futura‐ pico sensors

Mammalian and insect cells

Vessels with integrated optical DO and classical or optical pH sensors

Microbial cells Mammalian cells, insect cells, and stem cells Microbial cells

Finesse (Thermo Fisher Scientific)

0.5–2.2

Single‐use sensors for pH, DO, and temperature (SmartPack)

Mammalian cells, insect cells, plant cells, and microbial cells

Mobius® CellReady 3 l bioreactor

Merck Millipore

1–2.4

Vessels can be equipped with classical sensors (DO/pH)

Mammalian cells, insect cells, and stem cells

Ambr® (cell culture and fermentation)

Sartorius Stedim Biotech

0.01–0.25

Vessels equipped with optical sensor for DO and classical or optical sensor for pH measurements

Mammalian cells, insect cells, plant cells, stem cells, and microbial cells

0.6–2

Mammalian cells, insect cells, plant cells, and stem cells

CSRTM series

ABEC

50–4000

Bags can be configured with conventional or optical sensors

Mammalian cells

XcellerexTM XDR

GE Healthcare

4.5–2000

Reactors can be configured with conventional sensors

Mammalian cells, insect cells

XcellerexTM XDR‐MO fermenter

10–500

Microbial cells

Mobius® CellReady

Merck Millipore

10–2000

SensorReady technology for inserting of different sensors in a loop

Mammalian cells, insect cells, and stem cells

Allegro® STR

Pall Life Sciences

10–2000

Ports for aseptic insertion of conventional and/or single‐use sensors

Mammalian cells, insect cells

8–960

Bags can be configured with conventional sensors

PadReactor® BIOSTAT® STR

Sartorius Stedim Biotech

12.5–2000

Bags are equipped with single‐use sensors for DO and pH

Mammalian cells, insect cells, stem cells, and plant cells

HyPerforma S.U.B

Thermo Fisher Scientific

10–2000

Ports for integration of conventional or single‐use sensors for DO and pH

Mammalian cells, insect cells

HyPerforma S.U.F Wave‐mixed with flexible bag (1‐D motion)

Applications

SmartVesselTM

UniVessel® SU

Stirred with flexible bag

0.1–40

Instrumentation

6–300

Microbial cells

AppliFlex

Applikon

5–25

Bags can be configured with conventional sensors for DO/pH

Mammalian cells, plant cells

SmartRocker

Finesse (Thermo Fisher Scientific)

5–25

SmartPack sensors for DO, pH, and temperature

Mammalian cells, insect cells, and plant cells

ReadyToProcess WAVE 20/50

GE Healthcare

0.05–25

Bags are equipped with optical sensors for DO and pH

Mammalian cells, insect cells

WAVE Bioreactor 200

5–100

Xuri Cell Expansion System

0.3–25

BIOSTAT® RM

Sartorius Stedim Biotech

0.1–100

Human immune cells Optical bags are equipped with single‐use sensors for DO/pH

Mammalian cells, insect cells, stem cells, human immune cells, and microbial cells (Continued)

4  Single‐Use Bioreactors – An Overview

Table 4.1  (Continued)

Bioreactor type

Bioreactor brand

Vendor

Working volume (l)

Wave‐mixed with flexible bag (2‐D motion)

CELL‐tainer

BIOTECH BV

Wave‐mixed with flexible bag (3‐D motion)

Allegro® XRS 25

Orbitally shaken with flexible bag

OrbShake

Instrumentation

Applications

0.25–200

Bags are equipped with optical sensors for DO/pH; Possible to integrate glucose, lactate sensors

Mammalian cells, plant cells, and microbial cells

Pall Life Sciences

2–20

Bags contain optical sensors for DO and pH

Mammalian cells, insect cells

Kühner AG

3–200a

Bags are equipped with optical sensors for pH and DO

Mammalian cells, insect cells, human cells, and plant cells

The focus is only on selected wave‐mixed, stirred, and orbitally shaken versions. The order of the bioreactor types does not suggest the weighting factors. The table has no claim to be exhaustive. a  Prototype 2500 l.

Oscillatory bag movement φ = f(t)

42

1-D motion

2-D motion

3-D motion (biaxial motion)

t = ti

t = ti

t = ti

t = ti + ∆t

t = ti + ∆t

t = ti + ∆t

t = ti + n· ∆t

t = ti + n· ∆t

t = ti + n· ∆t

Figure 4.4  Schematic diagrams of wave‐mixed bioreactors with 1‐D, 2‐D, or 3‐D bag motion.

fluid dynamics (CFD) simulations, that the energy dissipation and shear stress pattern in wave‐mixed bioreactors are more homogeneous than in stirred cell culture bioreactors with Rushton turbines or paddle impellers. In general, the bubble‐free surface aeration takes place as the medium surface is continuously renewed. Oxygen transfer and its influence on the cultivation result have been investigated for most of the available systems [3]. The various wave‐mixed bioreactors differ  in their control mechanism, their bag design (film material, scale, and dimensions), the installed sensor types, and the kind of platform movement being

one‐dimensional (1‐D), two‐dimensional (2‐D), or three‐dimensional (3‐D) (Figure  4.4). The intensity of the mass and energy transfer and, therefore, cell growth and product expression, can be directly controlled through wave generation and propagation. However, to control these parameters sufficiently, a good system understanding is necessary. Thus, a characterization of the system with biochemical engineering methods prior to system usage is meaningful. Practical examples of biochemical engineering investigations of wave‐ mixed bioreactors can also be found in Chapter 22, part II of this textbook.

4.3  Comparison of the Current, Most Common SUB Types

Today, the 1‐D oscillatory mixing concept is most often applied (see Table  4.1). The fluid flow inside the 1‐D motion bags can be characterized by a modified Reynolds number (Remod) introduced by Eibl et al. [89], where turbulent conditions occur above a critical Reynolds number of 1000. As defined by the authors, Remod depends on the working volume, the width of the culture bag, the liquid level, the rocking rate, the kinematic viscosity of the fluid, and an empirical bag‐dependent constant. In contrast, Oncül et al. [90] used the nondimensional Womersley number (Wo) in combination with a parameter β to quantify the unsteady nature of the flow in wave‐mixed bioreactors. They stated that t­ urbulent conditions appear in oscillating wave‐mixed bioreactors when β exceeds 700 for Wo greater than 8.5. The wave propagation promotes bulk mixing and off‐bottom ­suspension of cells or microcarriers (mainly for vaccine production or stem cell expansion) in addition to surface aeration. Reported mixing times and oxygen mass ­transfer coefficients (kLa) in the range of 4–800 seconds, respectively, 5–50 per hour make wave bags with 1‐D motion suitable for most cell cultures, especially for those with low and medium oxygen demand and Newtonian fluid flow properties [3, 62, 89]. For 1‐D motion bags, it was demonstrated that increasing the rocking rate and angle is more effective in increasing the oxygen transfer than increasing the aeration rate. Limiting oxygen transfer can occur in the previously mentioned wave‐mixed bioreactors when fast‐growing plant cells or aerobic microorganisms are grown [11, 12, 15]. Eibl et  al. [87] showed that fast‐growing (doubling time ≈16 hours) tobacco Bright Yellow‐2 suspension cells were growth limited due to mass and gas transfer limitations in the bag. The cell growth was accompanied by a significant increase in viscosity from 0.001 to 0.41 Pa s, which resulted in poor mixing conditions and, therefore, in a reduction of the oxygen transfer. As exemplified by  Hitchcock [91] for bacterial human papilloma virus vaccine production involving high optical densities ­ (OD600  = 7–8), the high oxygen level required can be achieved by operating a BIOSTAT® RM with low culture volume (50 l bag with 5 l working volume). In contrast, Dreher et al. [92] showed that by using a linear feeding strategy based on a model for the oxygen consumption, an OD600 of 130 can be achieved for E. coli BL21 in the BIOSTAT® RM (10 l bag, 5 l working volume, 10° rocking angle, 42 rpm). Newer developments, such as the Allegro® XRS 25 or the CELL‐tainer, provide kLa‐values of up to 73 per hour respectively, 450 per hour by a 2‐D or 3‐D bag motion. The adapted bag motion creates a higher grade of turbulence in the system. Thus, these systems are suitable for cultures with high oxygen demands or with special mixing requirements, such as non‐Newtonian culture broths. The CELL‐tainer combines a vertical rocking

motion with a horizontal translation, whereas the XRS 25 bioreactor system comprises a 3‐D bag and uses a simultaneous biaxial rocking motion that gives the flow a tumbling characteristic. Westbrook et  al. [93] used the CELL‐tainer system successfully for the cultivation of E. coli BL21 and achieved comparable results as in stirred bioreactors. Furthermore, the linear scalability of the system was shown by Junne et al. [94] based on the maintenance of the main geometric features (the ratio of angle and plate width). They were able to keep the main fluid characteristics and shear forces similar over different scales. Due to the homogenous hydrodynamic stresses, even at the larger scale, the system seems to be suitable for the cultivation of shear‐sensitive algae or filamentous fungi, which are an attractive source for the production of food supplements and new antibiotics [14]. The Allegro® XRS 25 system provides cell culture performance comparable or even superior to mammalian (CHO and hybridoma) and insect suspension cells [95, 96]. It is undoubted that wave‐mixed bioreactors with 1‐D motion are best investigated and most often applied (compare Table  4.1 and also Chapter  22). CFD simulations revealed the presence of laminar flow conditions and very low shear stress levels of max. 0.01 Pa. These values are far below the values reported for stirred cell culture bioreactors and the critical range of anchorage‐ dependent growing cells (0.7 Pa), which can result in lethal damage. This encouraged different users [97–100] to use wave‐mixed 1‐D motion bags for expansion of hMSCs and human immune cells such as CAR T‐cells. In this context, GE Healthcare and Sartorius Stedim Biotech have launched special bags for processing primary cells. In addition, wave‐mixed 1‐D motion bags have proven themselves in animal cell‐based production processes of viruses for gene therapies, as well as for vaccines and for virus‐like particle vaccines [3, 6, 7, 101, 102]. Finally, secondary metabolites and recombinant proteins based on plant cell and tissue cultures are manufactured in wave‐mixed 1‐D motion bags [10, 103]. Prominent commercial product examples are Mibell Biochemistry’s PhytoCellTec Malus domestica and RootBioTec HO, and Greenovation’s moss‐a Gal (see also Chapter 28). It is worked with both plant cell suspension and hairy roots. But, despite the many different successful applications of wave‐mixed bioreactors with 1‐D motion, they are most frequently used in inoculum productions with feeding and since recently in perfusion mode (see also Chapter 14). 4.3.2  Stirred SUBs The experience from over 25 years of developing reusable stirred‐tank bioreactors for mammalian cells is reflected in the design of stirred SUBs. Nowadays, systems up to

43

44

4  Single‐Use Bioreactors – An Overview

4000 l working volume, with either rigid vessels (10 ml to 40 l working volume) or flexible bags (4.5–4000 l working volume) and height to diameter (H/D) ratios of 1.5:1– 2.2:1, are commercially available from different suppliers. The largest SUB with 4000 l has only recently been launched by ABEC and is currently implemented by Wuxi in Shanghai. The rigid systems are normally free‐standing, plastic vessels made of polycarbonate, which do not require a support container. Due to their rigid properties, they are not susceptible to folding stresses which, as with bag‐based systems, can lead to layer breakup and thus leakage. In the case of culture bags, it is important that they are properly installed and fit perfectly into their holding device to achieve optimum performance, especially in terms of heat transfer. Cavities, pockets, and folds, which can be attributed to the bag unfolding during installation, need to be prevented. At smaller scales, heating or cooling is performed by heating blocks, block coolers, or double jackets. Regardless of the scale, gas vent filter heaters are normally used for stirred SUBs instead of exhaust gas condensers. An exception is the BioBLU bioreactor (Eppendorf ), which uses a liquid‐free Peltier element that controls the temperature and condensation in the exhaust air. A similar principle is also used for the ambr 250 system (Sartorius Stedim Biotech) in order to reduce the loss of fluid by evaporation. The BioBLU® and the ambr 250 represent both stackable modular systems with up to 32 or more parallel cultivation vessels. The cultivation containers provide working volumes of 100– 250 ml and are suitable for screening experiments and process developments, allowing design of experiment (DoE) approaches and Quality by Design compliant proceedings. Both systems offer either a version for mammalian or microbial cells, while different types of impellers (mammalian  =  pitched blade impellers, microbial = Rushton impellers) are used to meet the different requirements of the production organisms. In addition to applications with mammalian (e.g. Chinese Hamster Ovary, CHO) and microbial (e.g. E. coli and Pichia Pastoris) cells, the mammalian versions of the two systems seem to be suitable for the processing of primary cells, especially for stem cells, due to the high degree of automation, the parallelization possibility, and the low initial working volumes. Dufey and his coworkers [104] described the successful use of the BioBLU® 0.3c system for the microcarrier‐based expansion of hMSCs. They were able to generate 1 × 108 cells in a working volume of 250 ml, while the cells retained their stem cell‐specific properties. Olmer et  al. [105] described the successful expansion of human pluripotent stem cells in the same system, where cell yields of up to 2.3 × 106 cells/ml were obtained in a seven‐day culture. Another popular stirred SUB system for screening studies or media optimization is the ambr 15, which provides working volumes of

between 10 and 15 ml with an automated liquid handling. Unlike the ambr 250 and other rigid SUB systems, the ambr 15 has a cubical shape and no internals, except the optical sensor spots for pH and Dissolved Oxygen (DO) measurement, the impeller, and the impeller shaft. Regardless of the unique vessel shape of the ambr 15, Nienow and his coworkers have shown that the ambr 15 is superior to conventional shake flasks, which are frequently used in early process development or for clone selection [106]. They showed that the performance of an IgG4‐expressing CHO cell line in the ambr 15 is more comparable to large‐scale bioreactors than in shake flasks. Rafiq et al. [107] used the ambr 15 successfully for the microcarrier‐based expansion of primary hMSCs from different donors in serum‐free culture medium and achieved comparable cell densities (up to 2.6 × 105 cells/ ml) as in conventional single‐use spinner flasks. The benchtop‐scale stirred SUBs, including the UniVessel® SU (Sartorius Stedim Biotech), the Mobius® CellReady 3L bioreactor (Merck Millipore), the BioBLU 1c and 3c (Eppendorf ), and the SmartVesselTM bioreactor (Finesse), are close to the classical design of reusable, stirred cell culture bioreactors. A combination of a modified Rushton turbine (bottom) and a three‐bladed segment impeller (top) is implemented in the SmartVesselTM bioreactor. All other mammalian versions of the single‐ use stirred bioreactors with a rigid vessel are equipped with one or two top‐mounted marine or segment blade impellers and enable specific power inputs in the range of 50–200 W/m3 to be reached, which are sufficiently high for common mammalian‐based processes. The Mobius® CellReady 3 l bioreactor, equipped with standard sensors and an Applikon ez‐Control process control unit, has been investigated by different groups. Detailed engineering parameters of this system can be found in [3] and in Chapter 22 of this textbook. Kaiser et al. have shown that similar cell densities and product activities can be achieved in the 3 l Mobius® CellReady in the same way as in stirred 3 l glass cell culture bioreactors. Furthermore, Cierpka et al. [108] and Lawson et al. [109] expanded hMSCs in 3 and 50 l Mobius® CellReady. Another well‐characterized single‐use system at the benchtop scale is the UniVessel® SU bioreactor, which was the first commercially available, rigid, SUB agitated by two‐stage impellers. The three elements of the segment blade impeller are similar in shape to the “Elephan ear” impellers of the BioBLU, but they have a lower blade angle of 30° (vs. 45°), which results in lower power inputs. The system has already been studied extensively by different groups in order to predict fluid flow, specific power inputs (0.4–435 W/m3), mixing times (3–100 seconds), oxygen transfer levels (10–50 per hour), and hydrodynamic stresses [3, 110, 111]. In addition to the countless applications with mammalian cells [112, 113],

4.3  Comparison of the Current, Most Common SUB Types

Figure 4.5  HyPerforma S.U.B 2000 L bioreactor. Source: Photo courtesy of Thermo Fisher Scientific Inc., Waltham, MA, USA.

recent investigations have shown the successful application of the UniVessel® SU for plant cell suspension cultures, such as Corylus avellana suspension cells. Schirmaier et  al. [114] used the UniVessel® SU for the expansion of hMSCs with microcarriers and achieved a peak cell density of 0.21 × 106  cells/ml with a serum‐ reduced culture medium. Based on these initial studies, Jossen et al. [20] found in a DoE-based CFD design study that the hMSC peak cell density in the UniVessel® SU can further be increased by 240% with an adapted impeller design (blade angle of 45°) and impeller position (offbottom clearance/vessel diameter ≈0.26). The single‐use stirred bag systems sold by Thermo Fisher Scientific (HyPerforma S.U.B, Figure 4.5) and GE Healthcare (Xcellerex) for use with animal cells and for volumes of up 2000 l have been proven as real alternatives to stirred bioreactors made from stainless steel. Both systems work with cylindrical bags in which low‐ shear impellers and commonly used aeration devices such as microspargers and open‐pipe spargers have been preinstalled. Relevant impeller (d/D of 1:3–1:2) and reactor geometries (H/D 1:1–2:1) of these bag bioreactors were replicated from their steel counterparts. The S.U.B and the XDR‐Disposable Stirred Tank Bioreactors possess an angular, axial flow impeller positioned off‐center. This construction, which eliminates the fluid vortex on the surface, thereby removing the need for baffles, is

often used in cell culture bioreactors at the pilot scale [3]. Whereas the XDR‐Disposable Stirred Tank Bioreactor runs with a magnetically coupled, bottom‐driven marine impeller, the S.U.B is top driven and incorporates a seal. The latter requires a more laborious installation of the bag in the support container before filling with medium and inoculation with cells. The S.U.B and the XDR‐ Disposable Stirred Tank Bioreactor are currently used in many mAb and vaccine productions, in which mammalian (e.g. hybridoma, CHO, Vero, and Baby Hamster Kidney) cells and insect cells are grown. Similarities in the engineering parameters, product quantities (middle to high cell densities and titers in g‐ range), and product quality with reusable stirred cell ­culture bioreactors have been achieved by different companies, such as Amgen, Baxter, Centocor, DSM Biologics, Sanofi‐Aventis, Lonza, and Roche. Furthermore, the interest in microbial versions of SUBs has resulted in the design of the S.U.F 300 and the XDR MO 500. The BIOSTAT® STR of Sartorius Stedim Biotech comes as close as possible to the classical configuration of a modern reusable cell culture bioreactor and the bioreactor series is available up to 2000 l. As an option, the bag is equipped with a sparger ring or a microsparger and two axial flow three‐blade‐segment impellers, or a combination of one axial flow three‐blade‐segment impeller and  one radial flow six‐blade‐segment impeller [115]. Homogeneous mixing in the bag is achieved by the centered stirring system. Schirmaier et al. [114] established a microcarrier‐based expansion process with hMSCs in a 50 l BIOSTAT® STR under serum‐reduced conditions and achieved the highest peak cell densities (0.31 × 105 cells/ml) reported in the specialized literature at that time. They observed comparable growth kinetics and metabolic production rates in the BIOSTAT® STR 50 l as in small‐scale disposable spinner flasks. Pall Life Sciences have launched with the Allegro® STR system, a bioreactor, which deviates from the classical designs. The cultivation container consists of a cube‐shaped bag, instead of a classical cylindrical vessel, and is available at four different scales (50, 200, 1000, and 2000 l). In contrast to the classical stirred SUBs, the support container of the Allegro® STR integrates notches on the sites of the bag which act as baffles. The bottom‐ mounted pitched‐blade impeller generates mixing times Foresee filtration of the resin to avoid polymer dust or other contaminants going with the resin in the extruder

Figure 8.20  Raw material arrives in bags.

Key points having to be considered: > Have a good procedure for the opening and the closing of the bags to avoid contaminating the resin > Have closed containers where the resin can be emptied in > Have adjusted garments for the operators emptying the bags > Do this preferably in a controlled environment

Cleanroom environment

Industrial environment

Bubble Roll of film Resin pellets

Air ring

Extruder

Die

Figure 8.21  Blown film extrusion cleanroom setup.

the tubing cannot be connected directly to the film. In more complex bags like disposable bioreactor bags, there can be more than 10 different fitments and components connected to the film or to each other to support the functionality of the bag. To make rigid and semi‐rigid components out of ther­ moplastic polymer materials, there are many different methods like rotational molding, blow molding and

thermo forming. But the most frequently used method is injection molding. Injection molding starts just like the film manufacturing with an extruder. The setup of the extruder is slightly different, however (see Figure 8.23). The polymer material is fed through the hopper into the extruder and the material will start melting while being moved forward in the cylinder. The difference now is that, instead of generating a constant pressure to push

103

104

8  Production of Disposable Bags

Cleanroom environment

Figure 8.22  Cast film extrusion cleanroom setup.

Industrial environment

Stationary plate mould Hopper

Heating element

Movable plate Rods Nozzle Clamping cylinder

Cylinder for screw

Screw Motor and gear for screw rotation

Plasticizing cylinder Injection unit

Clamping unit

Figure 8.23  Injection molding installation.

the molten plastic at a constant rate through a die, the molten material is collected at the end of the screw in front of a nozzle. When sufficient material is in front of the nozzle, the extruder screw will be pushed forward by a cylinder and the screw will push the molten plastic through the nozzle into the mold. The mold is typically a  two‐piece metal block that holds the cavity of the ­component that needs to be molded. The molten mass of plastic is pushed into this mold where the material will start cooling down and solidifies. If the material is sufficiently solid, the mold will open and the component can be taken out. Identical to the film extrusion process, extra attention has to be paid regarding the production of components used in biopharmaceutical manufacturing. The prepara­ tion of the raw material is already covered under the film extrusion section, and also the feed into the extruder is quite similar. During the injection molding of these components, release agents should be avoided, as they can generate extra contamination risks. To control the overall cleanli­ ness of the component leaving the mold, we can put the complete injection molding installation in a cleanroom, or we can only put the back end in a cleanroom and leave the injection part out, or only generate a local cleanroom

environment around the area where the mold opens. Whatever we choose, we need to guarantee that the component has no extra contamination risk coming from the outside from the time the mold opens, until the component is double packed. 8.3.4  Quality Insurance When dealing with disposable bags, several subcompo­ nents have to be considered like the film but also connec­ tors, caps, fitting, filters, and tubing. In this part of the chapter, we will focus on the material used in film and components directly in contact with the content of the bags. Components such as filters (see also Chapters 9 and 10) and tubing (see also Chapter  5) are covered in other parts of this book and we will not discuss them here. We will distinguish tests performed during qualifi­ cation, tests performed periodically, for every release of batches or continuously. This segregation corresponds to the one described in the Bio‐Process Systems Alliance (BPSA) guideline on component quality [2, 3]. Qualification tests are tests performed during material and/or process qualification. Periodic tests are tests done on a frequency lower than every batch but more than only initial qualification. Release tests are tests performed on

8.3  Film Manufacturing and Molding

every batch as a requirement for lot release. When ­dealing with new parts or components, specific tests are foreseen to qualify the material to confirm its suitability in the manufacturing of disposable bags. However, testing and validation requirements for film and components are very much linked to the application and should be established by carrying out a risk assess­ ment for the application. Product specification depends on the level of safety required for each step in which the single‐use systems will be used [4]. If disposable bags are made sterile by gamma irradiat­ ing them, the impact of gamma irradiation has to be assessed on the material. Indeed, gamma irradiation increases the cross linking inside the polymer. To test the impact of gamma irradiation, the qualification tests listed below needs to be performed on components gamma irradiated at the upper limit allowed during standard sterilization process (i.e. if standard irradiation is performed at min 25 kGy with a gamma irradiation range between 25 and 50 kGy, test must be performed after irradiation at a minimum of 50 kGy). This part does not plan to cover all existing tests but to review the tests that are relevant for standard applica­ tions based on the authors experience. As already men­ tioned, specific applications may imply other requirements based on risk analysis. For clarity, we have segregated tests by type covering mechanical tests, phys­ ical tests, biological tests, material dependent tests, extractable tests, chemical compatibility tests, functional tests, sterility requirements, contamination require­ ments, and expiry date. In the subsequent tables, we have listed the method commonly used by our company to characterize materi­ als. Other methods exist in the literature. For this reason, we have listed them as alternatives to make the reader aware of them. 8.3.4.1  Mechanical Tests

Table 8.1 lists the different methods used in mechanical tests to characterize film and components in disposable bags. 8.3.4.2  Physical Testing

Table 8.2 gives a list of typical physical tests mainly linked to permeability performed on films and components. Permeability tests are important to assess if the disposa­ ble bag can be used with product for example sensitive to oxygen. 8.3.4.3  Biological Testing

The biocompatibility requirements of a material depend on how it will be used within the process (e.g. media kitchen versus purification) and vicinity to the human body [2, 3]. Table 8.3 gives a list biological tests performed

on films and components. Those tests are ­performed to assess the toxicity of material used in disposable bags and impurities extracted from material. 8.3.4.4  Material‐Dependent Tests

In this part, we cover material specific tests listed in reg­ ulation, standard, or pharmacopeia. As already men­ tioned, material covered here are the one entering in composition of film and component in contact with the biocontainer content. Table  8.4 lists the material‐ dependent test to be performed for those materials. Most of the current standards deal with the use of dis­ posable biocontainers for drug delivery, and do not address containers in themselves. Only some of the exist­ ing standards are relevant, and must be adapted to the bag configuration [2, 3]. As a specific example EP 3.1.5 cannot be applied to a multi‐layer film since some of the layer are made in another polymer than EP. On the other hand, applying this monograph to the contact layer is also not perfectly correct since some requested proper­ ties (such as extraction) are given by the complete multi­ layer structure. 8.3.4.5  Extractables and Leachables

Polymers, used in the production of pharmaceutical con­ tainers and medical devices, cannot be considered pure compounds. Polymers should be seen as a blend of the base polymers with a broad range of chemicals that may be added for several reasons. Polymer additives are added to improve the processability of the polymer or to enhance its end‐use performance in various ways. Moreover, residues of unexpected and undesirable com­ pounds may also be present and can affect the biocom­ patibility and the toxicity of the materials [5]. Residues may originate from un‐reacted monomers, solvents used in the process, polymerization catalyst, surfactants, or polymer degradation products. The concentration of these chemicals/impurities in  the polymer can range from a few μg/kg to the ­p ercentage level and their migration behavior is often not known or is considered as proprietary informa­ tion by the supplier. This makes it very difficult to assess the toxicological potential of products that may migrate out of the polymers during contact with pharmaceuticals. Determination of extractables and leachables for sin­ gle‐use system, must be addressed as part of the process validation (see also Chapters 11, 17, 18). When properly evaluated, extractable and leachable can be addressed, and currently lot of efforts are made to address this early in the component selection process [6]. There is now a general consensus among the industry and regulatory agencies on the definitions for extractable and leachables (see Box 8.1).

105

106

8  Production of Disposable Bags

Table 8.1  Mechanical tests used to characterize film and components. Test type and general description

Test reference

Summary description of test reference

Gelbo flex Determines the flex resistance of materials by the formation of pinholes.

ASTM F392‐93

This test covers the flex crack resistance of materials by the formation of pinholes. Specimen (=sealed film tube) are twisted and horizontally folded at a constant rate and at different test cycles 500, 1000, 5000, or 10 000 test cycles. The number of pinholes is measured by filling the tube with H2O.

Dimension Determines dimension of the film or component.

In‐house

For thickness, the film or component is placed in a calibrated digital system that measures the thickness of the film. Thicknesses of the different layers of a multilayer film are evaluated by optical microscopy. Other dimensions can be evaluated by measuring device and/or caliber.

Tensile strength and elongation A measure of the force required to stretch a material to its breaking point [2].

ASTM D882‐91 Alternative method: ISO 527‐3

Used for thin sheets of plastic, including film, less than 1.0 mm (0.04 in.) thick. Performed at 23 ± 2 °C at RH of 50% ± 5%. Speeds are determined by the material in question and a reference table provided by the standard. One type of specimen design [2].

Puncture resistance Puncture resistance testing predicts the durability of the film while in use. Films with high puncture resistance correspond with material that can absorb the energy of an impact by both resistance to deformation and increased elongation. Puncture resistance, measured in energy units, evaluates the film strength and extensibility properties [2].

FTMS 101B   Alternative method: ISO 7765‐2

A pressing bar with a fixed diameter will be pushed through the film at a constant speed. The pressure force, needed to break the film, indicates the puncture resistance of the film. Test conditions: temperature 23 ± 2 °C and RH of 50 ± 5%

Dart impact Test method covers the determination of the energy as part of mechanical properties that causes plastic film to fail under specified conditions of impact of a free‐falling dart.

ASTM D1709‐01A

Two testing methods depending on the size of the striker, which are determined by the impact resistance of the material, are described. The standard technique is a staircase method to drop a weight and increase or decrease depending in pass or fail. Standard apparatus and striker has been described. Test conditions are 23 °C ± 2 °C and 50% RH ± 5%. [2]

Tear resistance Determines the tear resistance of flexible plastic film and sheeting at very low rates of loading [2].

ISO 6383/1 Alternative method: ASTM D1004

Test is designed to measure the force to initiate tearing.

It is important to note that not all leachables may be found during the extractables study. Some buffer or components of the drug formulation may interact with the polymer or one of the additive to form a new “leach­ able” contaminant that was not identified during the extractable study. New “leachables” can also be found if the processing conditions are more severe than the one used for the extractable/leachable study or if the analyti­ cal method used are different [3]. The relationship between extractables and leachables can be illustrated by the Venn diagram shown in Figure  8.24. Leachables include known extractables as well as those that are chemically modified by drug for­ mulation or processing conditions. There are no specific standards or guidance referencing extractables and

leachables from disposable bags, but USP is a good guidance to be used for the design of a meaningful extraction study. Many references that do apply were written to address all processing materials and equip­ ment without focusing on materials of construction. However those references are sufficiently broad to include leachables [6]. The challenge of an extractable and leachables study is to give maximum information to users of disposable biocontainers in order to identify if further information or testing is required for their spe­ cific application. Since extraction studies require exaggerated condi­ tions, samples may be exposed to aggressive solvents (i.e. dichloromethane [DCM]), but less stringent approaches are most of the cases used. The following analyses are

8.3  Film Manufacturing and Molding

Table 8.2  Physical tests used to characterize film and components. Test type and general description

Test reference

Summary description of test reference

O2 transmission rate Determines the steady‐state rate of transmission of O2 gas through material.

ASTM D3985     Alternative methods: ASTM F1927, ASTM D3985/D1434, ISO 15105

Test method to evaluate steady‐state rate O2GTR. The specimen is to be conditioned in a dry environment less than 1% RH. Then it separates two chambers at ambient pressure. The chambers are then purged – one with N2 and the other with O2. O2 permeation is measured with a coulometric detector. No operating condition is required. Typical conditions: 23 °C, 90% RH inside; 50% RH outside

CO2 transmission rate Determines the steady‐state rate of transmission of CO2 gasses through material.

In‐house method 23 °C, 0% RH

Water vapor transmission rate

ASTM F1249   Alternative methods: ISO 15106

Uses a modulated infrared sensor to measure water vapor permeability at 5 and 23 °C with 0% RH on the outside and 100% RH on the inside to simulate worst‐case use conditions for a fluid storage container. This is a diffusion cell procedure with a dry gas chamber with controlled temperature and RH and a wet cell controlling the same variables. The WVTR is calculated from an infrared pressure sensor that compares WVTR to known value.

Table 8.3  Biological tests used to characterize film and components. Test type and general description

Frequency

Test reference

Summary description of test reference

Biological reactivity – in vitro Evaluates the response of mammalian cell cultures to extracts of polymeric materials.

Qualification

USP Alternative method: ISO 10993‐5

Extracts are obtained by placing the test and control materials in separate cell culture media under standard conditions. Cells are observed for visible signs of toxicity (such as change in size or appearance of cellular components or a disruption in their configuration) in response to the test and control materials [2].

Biological reactivity – in vivo Evaluates the response in animals to exposure of polymeric materials.

Qualification

USP   Alternative method: ISO 10993‐6,10, and 11

USP Biological Reactivity Test – in vivo for Class VI Plastics, is a series of three tests: systemic toxicity, intracutaneous reactivity, and implantation. The first two are designed to determine the systemic and local, respectively, biological responses of animals to plastics and other polymers by the single dose injection of specific extracts prepared from a sample. The implantation test is designed to evaluate the reaction of living tissue to the plastic and other polymers by the implantation of the sample itself into animal tissue [2].

performed on sample extract: volatile organic c­ ompounds (VOC, analytical technique such as Headspace  – GC/ MS), semi‐volatile organics (SVOC, analytical tech­ nique:  GC/MS) and non‐volatile organic compounds ­(analytical technique, LC/MS). Leachable studies are more  representative in case of the use of disposable biocontainers. In order to be relevant, the study should be performed on samples mimicking mixing bags and including film and subassemblies including for example fitment and tubing. Samples are submitted to different solutions usu­ ally under accelerated conditions (40 ± 2°C and 75%

humidity) except for some solvent (i.e. dimethyl sulfox­ ide, DMSO). The supplier tries to cover all types of appli­ cation with representative products [7, 8]. Examples are water for injection, 50% ethanol in water, 0.1 M phos­ phoric acid, pH 3 solution (0.001 N HCl with 50 mM NaCl), pH 10 phosphate solution, 0.5 N sodium hydrox­ ide, or surfactant (e.g. polysorbate 80). The study covers a time period that should include most of the standard applications (i.e. up to six months). Analyses performed on every solution and time point are aimed at: pH‐conductivity, organic compounds (analyti­ cal technique: TOC), volatile organic compounds (VOC,

107

108

8  Production of Disposable Bags

Table 8.4  Material dependent tests used for film and/or components in disposable bags. Test type and general description

Frequency

Test reference

Summary description of test reference

Tests on plastic materials and components used to package medical articles Evaluates the physical and chemical properties of plastics and their extracts.

Qualification

USP

Test for PE, PP, PET. Measures the properties of impurities extracted from plastics when extracted in a model solvent (such as purified water and isopropanol) over a specified period and temperature. Includes the following: heavy metals, buffer capacity, and non‐volatile residue.

Tests on polyethylene with additives for containers for parenteral preparations and for ophthalmic preparations

Qualification

EP 3.1.5

Tests specific for PE including physical properties, extractable and impurities.

Tests on plastic containers for aqueous solutions for parenteral infusion

Qualification

EP 3.2.2.1

UPW is placed in a container and then extracted. Appearance of solution, acidity or alkalinity, absorbance, reducing substances, and transparency are evaluated.

Test on olefins in contact with food

Qualification

21 CFR 177.1520

Tests on PE and PP.

Tests on EVA in contact with food

Qualification

21 CFR 177.1350

Tests on EVA copolymers

Box 8.1  Definitions [3] of Extractables and Leachables Accepted by Industry and Regulatory Agencies Extractables: Chemical compounds that migrate from elastomeric, plastic or coating components when exposed to an appropriate solvent under exaggerated conditions of time and temperature. Leachables: Chemical compounds, typically a subset of extractables, that migrate from elastomeric, plastic or coating components as a result of direct contact with the drug formulation under normal process conditions or accelerated storage.

analytical technique: Headspace  –  GC/MS), semi‐ and non‐volatile organics (analytical technique: GC/MS), and metals (analytical technique: inductively coupled plasma, ICP). 8.3.4.6  Chemical Compatibility Tests

Disposable bags have the potential risk to interact with the solution introduced in the bag. The aim of the chemi­ cal compatibility study is to test the resistance/compati­ bility of material exposed to defined solutions. Very few standard tests exist for this purpose. An adaptation of American Society for Testing and Materials (ASTM) D543‐06 “Standard Practices for Evaluation of the resis­ tance of Plastics to Chemicals” can be used to “grade” the compatibility of the film [2, 3]. The type of conditioning depends on the end‐use of the material. Change in weight, dimensions, color, surface

Extractables Leachables

Figure 8.24  The relationship between extractables and leachables. Source: Adapted after Bio‐Process System Alliance [3].

quality, appearance and strength properties are reported before and after contact. A general compatibility status is given upon testing results classified between “excellent compatibility” and “not recommended use” [2, 3]. Disposable bag manufacturers aim to cover a broad range of solvent and process conditions in initial tests. Preferred tests are given in Table 8.5. If a disposable bag has to be used with another solu­ tion, the user may either correlate the solution to one of the tested solution (if the solvent is similar, the compatibility can be considered as similar) or conduct a specific study [2, 3]. 8.3.4.7  Functional Tests – Assembly Test

The aim of functional tests is to prove that disposable bags withstand the requirements of standard use. The type of test performed is test of strength of assembly but also integrity test after opening and closing of ports. Test of strength in assembly is done to validate the assembly of component. The test performed is a “pull off” test, which measures the requisite force to separate components

8.3  Film Manufacturing and Molding

Table 8.5  Example of solutions for chemical compatibility test. Product

Temperature (°C)

Contact time

 1

Acetic acid

60

3 h

 2

NaOH 50%

60

3 h

 3

MgSO4·7H2O

30

7 d

 4

MnSO4·7H2O

30

7 d

 5

KH2PO4

30

7 d

 6

Ethanol 55%

30

7 d

 7

Eau de Javel (NaClO) 15% + NaOH 0.5 M

50

3 h

 8

Riboflavine (vitamin B2)

30

7 d

 9

Kanamycine sulfate

30

7 d

10

Caseine hydrolysate

30

7 d

11

TRIS: Tromethanine (HOCH2)3CNO2

30

7 d

12

Diethanolamine (DEA/Ureum/SLS)

30

7 d

13

Guanidine HCl

30

7 d

14

Yeast extract

30

7 d

Figure 8.25  Lloyd LRX force tester, setup for defining the pull‐off force of tubing on fitment.

from an assembly. Figure 8.25 shows one example of such a test for an assembly of tubing to a fitment. Another type of test consists in applying an integrity test (see Section  8.4.2) to the bag after an action ­corresponding to normal use of the bag such as for example opening and closing of fitment/connector. The test is considered as successful if no loss of ­integrity is observed [9]. 8.3.4.8  Functional Test – Differential Scanning Calorimetry (DSC)

DSC allows the study of thermal transition of polymer. Thermograms obtained by DSC allow characterization

of polymers (Figure  8.26) [10]. Typical information is melting temperature (Tm), glass transition (Tg), and other thermal transitions (such as Brittle temperatures). The Tg, is the temperature at which the amorphous phase of the polymer is converted between rubbery and glassy states. The brittle temperature is defined as the temperature at which the material becomes brittle. This information is an important mechanical property. Some authors such as Barbaroux and Sette [11] reported that the correlation between the brittle temperature, the glass transition temperature, and the mechanical behavior at cold temperature is not obvious especially in the case of multi‐layer structure. DSC can also give information on the molecular arrangements of polymers, e.g. differentiate between LDPE, LLDPE, HDPE different nylons, homo‐ and co‐polymers, and polymer blends. It can be used to differentiate points linked to multi‐layer laminates. DSC can also be used to determine the thermal stability of poly­ mers especially to oxidation. 8.3.4.9  Sterility Tests

If the application requires a sterile single‐use system, the sterilization steps are performed using gamma irra­ diation. A validation of the minimum necessary dose to claim the product sterile at 10−6 sterility assurance level (SAL) has to be performed following International Organization for Standardization (ISO) 11137 stand­ ard. Different methods are presented, all are based on the estimation of the initial bioburden contamination (see release test ISO 11737). Due to size of samples, the sterilization dose method VDmax is usually used for biocontainers.

109

8  Production of Disposable Bags 61.0 60.5

104.95°C 59.4352 mW

60.0 Heat Flow Endo up (mW)

110

116.14°C 80.5116 mW

148048 153779

59.5 59.0 58.5 58.0

184.06°C 56.7663 mW

57.5

209.83°C 57.3767 mW

217.03°C 57.8092 mW

57.0 56.5 56.0 55.5 55.0

60

80

1) 2) 3) 4) 5) 6)

100

120

140 160 Temperature (°C)

Heat from 50.00°C to 260.00°C at 20.00°C/mh Hold for 2.0 mh at 260.00°C Cool from 260.00°C to 50.00°C at 20.00°C/mh Hold for 2.0 mh at 50.00°C Heat from 50.00°C to 260.00°C at 20.00°C/mh Hold for 2.0 mh at 260.00°C

180

200

220

240

7) Cool from 260.00°C to 115.00°C at 20.00°C/mh 8) Hold for 30.0 mh at 115.00°C 9) Cool from 115.00°C to 50.00°C at 20.00°C/mh 10) Hold for 2.0 mh at 50.00°C 11) Heat from 50.00°C to 200.00°C at 20.00°C/mh

Figure 8.26  DSC of a polymer.

8.3.4.10  Contamination Requirements

Contaminants include particulates, bioburden and endo­ toxins. For particulates and bioburden, the contamina­ tion requirements are linked to the application. Contamination will be less stringent in a process if a sterilization step or a filtration/purification step is used. Particulate contamination is covered in the release tests (Section 8.4.2). The approach taken is similar to one used in cleaning of a bioreactor. The amount a particulate pre­ sent in bag should be low enough to give a product com­ pliant with the United State Pharmacopeia (USP) . There is also a clear requirement for the bag to essen­ tially free of visible particulates and to have a better understanding of the particulate types. This is proven by checking compliance on the minimum volume that could be used in the vessel. Bioburden has been covered in the sterility requirements and are also covered in the release tests (Section 8.4.2). Regarding endotoxin, regardless of the application, endotoxin requirements must be met. This test is covered in the release test (Section 8.4.2). 8.3.4.11  Expiry Date

Product aging should be assessed because polymeric materials are known to age and their physical and com­ patibility properties may change with time [10]. Accelerated aging studies are frequently performed in accordance to ASTM F1980. For sterile single‐use sys­ tems, the study should also cover gamma irradiation.

Long term study allows the confirmation of the acceler­ ated study results and the assessment of tests that are not feasible on accelerated samples (e.g. sterility test).

8.4 ­Bag Manufacturing 8.4.1  Most Important Manufacturing Processes Used in the Production of Disposable Bags In the final production of disposable bags, all the differ­ ent components are joined together to create the final bag assembly. The most important joining technique used for the production of disposable bags is welding. Other joining techniques are mechanical fastening, which is typically used to connect tubing sets to the fit­ ments welded on the bag, or chemical bonding, which is mainly used to join rigid components which are not ther­ moplastic [12, 13]. Looking at the wide variety of bags used with different shapes and different fitments manufactured using differ­ ent joining techniques, we will only focus on the most important shapes and mention the most frequently used welding techniques (see also Chapter 5). Bag manufacturing can be split out in welding of film to create the bag chamber, welding different fitments or connectors into the bag chamber and then connecting

8.4  Bag Manufacturing

Figure 8.27  Two‐dimensional bag design.

the different functional elements (connectors, filters, clamps, valves, sampling bags, etc.) on the fitments through tubing sets. For the manufacturing of bag cham­ bers an important factor is the geometry of the bag. As described in Chapter 2, this geometry can be two‐dimen­ sional, two sheets of film are welded together to create a kind of pillow shaped bag when filled, or three‐dimen­ sional, where multiple sheets of film are welded together to create a cubical or cylindrical shaped bag when filled. The standard design of two‐dimensional bags is to have a front and back film which are welded to each other around the complete circumference of the bag with a holding mechanism at the top and the connected tub­ ing sets for filling, emptying, or sampling in the bottom sandwiched at the edge of the bag between the two films as shown in Figure 8.27. The way the tubing sets are connected to the bag can differ depending on the type of film material used. If the film is made out of flexible and thick materials, like ULDPE, PP, EVA, or PVC of 200 μm, and thicker, and the tubing is made out of a material that is compatible to the film material from a welding perspective, then the tubing can be welded directly in between the two films. If the film is more rigid, which is often the case in multilayer structures containing EVOH, PA or PET, or if the film material is incompatible with the tubing material from a welding perspective, then an intermediate fitment will be used. This fitment will have typically a boat shaped base structure, to facilitate the welding process, and a number of fitments where the tubing sets can be connected on (see Figure 8.27). Another alternative is to have a flange‐ based tubing fitment that is not welded in between the films at the edge of the bag but which is welded in the surface of the film. This type of bag design can generate some issues when the bag needs to be emptied. The standard design of three‐dimensional bags is a cubical shape. In some cases, cylindrical designs are made. In theory, there are many options to weld different

films together to generate a cubical bag. The level of automation usually determines what bag assembly method will be used. Fitments are always welded in the surface of the film using flange‐based fitments. The welding technique depends on the type of materi­ als used for the different components and their form. The minimum requirements to weld materials together are that they are thermoplastic and compatible. To weld flexible film based components together, the most fre­ quently used welding methods are radio‐frequency (RF) welding, constant heat welding, impulse welding, ultra­ sonic welding or laser welding. Except for laser welding, the films that need to be welded together are put in between the tools of the welding system and pressed together. In the case of RF welding, a rapidly reversing electric field between the upper and lower tool makes the polar molecules within the material start moving and due to the resulting internal friction generates heat, which makes the film melt and weld together. For radio‐fre­ quency welding, only polar materials can be used like PVC or EVA. This is a popular welding method for two‐ dimensional bags. With constant heat welding, the tools are heated up to a constant temperature using heating elements in the tools. The hot upper and lower tool are pressed together, melting the film materials and welding them together. Although a simple way of welding films, the downside is that heat has to go from the outside through the film to the welding interface. The film becomes then very elastic when the tools are opening again and can fracture when any film tension exists. That is why this method is typi­ cally used in multi‐layer structures where the outside layer, having a higher melt temperature than the contact weld layer, can maintain the strength of the film when the welding tools are opening again. With impulse welding, the tools contain a resistor wire which is pressed onto the film materials. Electrical current

111

112

8  Production of Disposable Bags

is sent through the wire, heating it up, melting the films and welding them together. Then, while still being pressed together, the current is stopped and the wire cools down. The tools are kept closed until the film material has cooled and sufficiently solidified. The welding principle is the same as for constant heat welding with the advantage that there is more freedom in the type of films that can be welded together due to the cooling function of this method. With ultrasonic welding, the lower tool is fixed and the upper tool is pressed on the lower tool with the films in  between; it then starts vibrating at a high frequency (20–40 kHz). The vibrations generate heat at the weld interface, welding the films together. In contrast, with laser welding, a high intensity laser beam is used to increase the temperature at the interface of the films, making them melt and weld together. To weld flexible film and a rigid component together, the same welding methods can be used as with welding films together. The main difference is that with methods that are transferring the heat through the material ­(constant heat and impulse welding) the heat will be ­generated by the tool that will get in contact with the film. To weld rigid components together, the most fre­ quently used welding methods are ultrasonic welding, and hot plate or mirror welding. With ultrasonic welding, the technique is as described for welding films together. With hot plate or mirror welding, a hot tool is kept in between the two surfaces that need to be welded. Then, these surfaces are pressed against the hot tool until they are starting to melt. Then, they are removed from the hot tool and pushed against each other. Naturally, this must happen fast enough to make sure the two surfaces are still in a molten phase when they are pressed together. All these different manufacturing steps need to hap­ pen in a clean room environment; thus, all equipment needs to be designed to meet cleanroom requirements. 8.4.2  Quality Insurance 8.4.2.1  Control of Incoming Material

Table  8.6 lists typical tests for incoming inspection of material entering in disposable bags. Some of these techniques are already used in the phar­ maceutical industry, for example Fourier transform infra­ red spectroscopy (FTIR), which can determine the type of polymer used in a film or a component. The infrared (IR) absorption spectrum exhibits narrow, closely spaced absorption peaks resulting from t transitions among the various vibration quantum levels [14]. The number of individual ways a molecule can vibrate is largely related to the number of atoms and thus the number of bonds it con­ tains. IR spectra give qualitative information regarding the

chemical species contained in a polymer. In multi‐layer polymers such as the ones used in disposable bags, the spectrum is complex. However, it is possible to confirm the different layer and to confirm the good positioning of the film in the manufacture of bag (Figures  8.28 and 8.29). For resin, other properties to be checked during incoming inspection are melt flow rate, density and melting point. Melting point can be measured using DSC (see Section 8.3.1). On components, such as fitment and connector, ­critical dimensions must be also controlled during the incoming inspection. This can be done using appropriate measuring device (care should be given to tolerance of this device) or calipers (Figure 8.30). Calipers give a pass/fail assessment but are easier to control multiple samples. 8.4.2.2  Release of Disposable Bags

Table 8.7 lists typical tests for release of disposable bags. 8.4.2.2.1  Integrity Tests

Integrity tests (see also Chapter 2) are employed to ver­ ify biocontainers, and fully, closed assemblies are fully sealed and integral. One important point is that the integrity test is intended to be performed on the bio­ containers and not on individual components (e.g. fit­ ments). Bags are inflated and the pressure in the bag is measured constantly (Figure 8.28). The change in pres­ sure is measured either inside the package itself or out­ side in a sealed package test chamber. The change in test chamber pressure can be determined by absolute pressure measurement, or it can be determined using a differential pressure manometer in which the test chamber pressure change is compared with of a control package. Several gasses can be used such as air, nitro­ gen (when stored products are sensitive to O2), and helium, for example [2]. In case of loss of integrity (i.e. holes or open seal), the pressure drops or the gas used to inflate the bag is detected. The absence of pressure drop or detection of no gas during a defined time is proof of the integrity of the bag (Figure 8.30). The more recent application of Helium Integrity test (HIT) for leak testing of biocontainers represents an advancement for the biotech industry, enabling the detec­ tion of very small leaks (e.g. 6 log of mammalian parvovirus over a continuous process including two interruptions of 13.5 hours each [30]. 9.4.3.4 Chromatography

Virus‐removal claims based on continuous chromatog­ raphy processes are predicted to be an inherent property of the resin itself. Several studies have shown that aspects of continuous chromatography such as frequent cycling

Acid

Product Base

Mixer 1

Acid

Mixer 2

Base

Inactivated product

Inactivated product

Figure 9.6  Working principle of disposable continuous virus inactivation strategies using in‐process mixing and hold (left) with product collection in mixer 1 and inactivation in mixer 2, and tubular plug‐flow reactors (right). Source: Courtesy of Pall Biotech.

123

124

9  Single‐Use Downstream Processing for Biopharmaceuticals

of columns, increased resin age, or the continuous flow schematic does not affect the virus clearance compared to batch processing [22]. To take process differences such as the increased binding capacity used in continu­ ous chromatography into account, it is recommended to adapt the virus‐clearance study design [22, 31]. 9.4.4 Formulation Performing ultrafiltration and diafiltration steps in con­ tinuous‐flow mode requires overcoming one of the main disadvantages of traditional TFF systems which is the repetitive membrane passage through the recirculation loop. Alternatives are provided by single‐pass TFF sys­ tems (SPTFF) which allow a continuous flow of liquid with only one flow‐ratio‐controlled module and pump passage. This SPTFF design eliminates the recirculation loop and hold container which allows a continuous flow and at the same time avoids potential foaming or mixing issues. Furthermore, the single‐pass design reduces the feed‐flow rate by factor 5–10 which results in smaller line sizes and thus smaller holdup volumes. Shear stress and residence times are furthermore reduced compared to conventional TFF [32]. For concentration, two design principles are available in single‐use format: (i) specifically designed modules with staged flow‐paths (Pall Biotech) and (ii) standard TFF cassettes operated at lower flow rates for higher feed‐to‐permeate conversion rates (Merck Millipore). Modules for continuous concentration are available from 0.065 to 3.5 m2 [32]. Diafiltration can be achieved by staged flow‐paths consisting of repetitive concentration and dilution. It has to be considered that the low linear velocities generally applied in continuous formulation can increase the risk of aggregation and membrane fouling. Further details on single‐use formulation and fill are given in Chapter 25 of this book.

9.5 ­Integrated and Continuous DSP The benefits of single‐use continuous processing tech­ nologies are enhanced if integrated into a hybrid or fully continuous DSP platform. Several fully integrated con­ tinuous DSP platforms have been developed in the past years by Merck & Co. Inc. in Kenilworth, USA [26], Sanofi‐Genzyme in Framingham, USA [33, 34], Bayer

in Leverkusen, Germany [35], and Pall Biotech in Westborough, USA [8]. A case study from the continuous DSP laboratory at Pall Biotech shows a fully single‐use end‐to‐end platform for mAbs. The platform starts with a primary clarification through acoustic separation fol­ lowed by continuous filtration and a pre‐concentration with inline concentration modules. Capture is performed with a multicolumn protein A chromatography with con­ secutive low‐pH virus inactivation using the in‐process mixing and hold technique. Polishing with an anion exchange flow‐through chromatography and a mixed‐ mode cation exchange is followed by continuous virus filtration, SPTFF for pre‐concentration, continuous dia­ filtration, final SPTFF concentration, and sterile filtra­ tion. The platform has shown to offer up to 74–95% savings in resin volumes across three different chroma­ tography steps, combined with 45% savings in buffer con­ sumption over the entire platform [8]. In addition to fully integrated DSP platforms, the implementation of continuous technologies in hybrid platforms has shown to provide processing benefits in several studies. This includes, for example, the combina­ tion of inline concentration modules with continuous capture chromatography for antibody purification [36]. The combination has shown to increase the specific pro­ ductivity of the chromatography by a factor of almost 4, from 8 to 31 g/l/h. Other examples show the potential of applying only one continuous step within a batch plat­ form such as a capture chromatography step [25]. Further examples on integrated continuous DSP can be found in Chapter 24.

9.6 ­Summary and Conclusions Single‐use technology in DSP is widely accepted in the industry from process development to larger scale and has shown to provide a range of benefits including sig­ nificant cost reduction. The current focus on process intensification in DSP supports the implementation of single‐use continuous processing technologies. The interconnection of continuous DSP unit operations pro­ vides process optimization possibilities throughout the platform as demonstrated by several adopters. However, the decision for single‐use or continuous technology has to be made on a case‐to‐case basis. Further develop­ ments toward integrated single‐use DSP are expected and offer further potential for process optimization.

Nomenclature ATF Alternating tangential flow filtration BioSMB Multicolumn simulated moving bed

DSP Downstream processing ICB Integrated continuous bioprocessing

­  References

ICH

International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use mAb Monoclonal antibody MCSGP Multicolumn countercurrent solvent gradient purification

PCC

Periodic countercurrent simulated moving bed SPTFF Single‐pass TFF systems SS Stainless steel SU Single‐use TFF Tangential flow filtration

­References 1 Rogge, P., Müller, D., and Schmidt, S.R. (2015). The

single‐use or stainless steel decision process: a CDM perspective. BioProcess Int. 13: 10–15. 2 Jacquemart, R., Vandersluis, M., Zhao, M. et al. (2016). A single‐use strategy to enable manufacturing of affordable biologics. Comput. Struct. Biotechnol. J. 14: 309–316. 3 Pollard, D., Brower, M., Abe, Y., et al. (2016). Standardized economic cost modeling for next‐ generation mab production. http://www.bioprocessintl. com/business/economics/standardized‐economic‐cost‐ modeling‐next‐generation‐mab‐production (accessed 15 September 2018). 4 Levine, H.L., Stock, R., Lilja, J., et al. (2013). Single‐ use technology and modular construction. http:// www.bioprocessintl.com/upstream‐processing/ upstream‐single‐use‐technologies/single‐use‐ technology‐and‐modular‐construction‐341774 (accessed 15 September 2018). 5 Whitford, W.G. (2010). Single‐use system as principal components in bioproduction. http://www. bioprocessintl.com/upstream‐processing/upstream‐ single‐use‐technologies/single‐use‐systems‐as‐ principal‐components‐in‐bioproduction‐307214 (accessed 15 September 2018). 6 Pollock, J., Coffman, J., Ho, S.V., and Farid, S.S. (2017). Integrated continuous bioprocessing: economic, operational, and environmental feasibility for clinical and commercial antibody manufacture. Biotechnol. Prog. 33 (4): 854–866. 7 Walther, J., Godawat, R., Hwang, C. et al. (2015). The business impact of an integrated continuous biomanufacturing platform for recombinant protein production. J. Biotechnol. 213: 3–12. 8 Gjoka, X., Gantier, R., and Schofield, M. (2017). Transfer of a three step mAb chromatography process from batch to continuous: optimizing productivity to minimize consumable requirements. J. Biotechnol. 242: 11–18. 9 Ötes, O., Flato, H., Winderl, J. et al. (2017). Feasibility of using continuous chromatography in downstream processing: comparison of costs and product quality for a hybrid process vs. a conventional batch process. J. Biotechnol. 259: 213–220.

10 Hummel, J., Pagkaliwangan, M., Gjoka, X. et al.

11

12

13

14

15

16

17

(2018). Modeling the downstream processing of monoclonal antibodies reveals cost advantages for continuous methods for a broad range of manufacturing scales. Biotechnol. J. https://doi. org/10.1002/biot.201700665. Konstantinov, K. and Cooney, C. (2014). White paper on continuous bioprocessing. https://onlinelibrary. wiley.com/doi/abs/10.1002/jps.24268 (accessed 15 September 2018). Bisschops, M. and Brower, M. (2013). The impact of continuous multicolumn chromatography on biomanufacturing efficiency. Pharm. Bioprocess. 1 (4): 361–372. Ko‐Hsu, F. and Bhatia, R. (2012). Evaluation of single‐ use fluidized bed centrifuge system for mammalian cell harvesting. http://www.biopharminternational.com/ evaluation‐single‐use‐fluidized‐bed‐centrifuge‐system‐ mammalian‐cell‐harvesting?id=&sk=&date=&pageID=2 (accessed 15 September 2018). Collins, M. and Levison, P. (2016). Development of a high‐performance integrated and disposable clarification solution for continuous bioprocessing. http://www.bioprocessintl.com/downstream‐ processing/separation‐purification/high‐performance‐ integrated‐and‐disposable‐clarification‐solution (accessed 15 September 2018). Grier, S. and Yakubi, S. (2016). Prepacked chromatography columns: evaluation for use in pilot and large‐scale bioprocessing. http://www.bioprocessintl. com/downstream‐processing/chromatography/ prepacked‐chromatography‐columns‐evaluation‐for‐ use‐in‐pilot‐and‐large‐scale‐bioprocessing (accessed 15 September 2018). Repligen. (2018). Scale‐up with opus: introducing opus 80 R columns, 2018. http://www.repligen.com (accessed 4 August 2018). International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use. (1999). ICH Q5A(R1): Viral Safety Evaluation of Biotechnology Products Derived from Cell Lines of Human or Animal Origin. https://www.ich.org/fileadmin/ Public_Web_Site/ICH_Products/Guidelines/Quality/

125

126

9  Single‐Use Downstream Processing for Biopharmaceuticals

18

19

20

21

22

23

24

25

26

Q5A_R1/Step4/Q5A_R1__Guideline.pdf (accessed 22 February 2019) Zhou, J. and Dehghani, H. (2007). Development of viral clearance strategies for large‐scale monoclonal antibody production. In: Advances in Large Scale Biomanufacturing and Scale‐Up Production (ed. E.S. Langer), 1–28. Rockville: ASM. Miesegaes, G., Lute, S., and Brorson, K. (2010). Analysis of viral clearance unit operations for monoclonal antibodies. Biotechnol. Bioeng. 106 (2): 238–246. Pall Biotech. (2018). Cadence virus inactivation system: automated semi‐continuous low pH virus‐inactivation on a single‐use mixing platform. https://biotech.pall. com/en/news/2017‐09‐25‐virus‐inactivation.html (accessed 15 September 2018). Bonham‐Carter, J. and Shevitz, J. (2011). A brief history of perfusion biomanufacturing. Bioprocess Int. 9: 24–30. Johnson, S., Brown, M., Lute, S., and Brorson, K. (2017). Adapting viral safety assurance strategies to continuous processing of biological products. Biotechnol. Bioeng. 1: 21–32. Jornitz, M. and Meltzer, T. (2006). Grow‐through and penetration of the 0.2/0.22 “sterilizing” membranes. https://www.researchgate.net/publication/290035973_ Grow‐through_and_penetration_of_the_02022_ sterilizing_membranes (accessed 15 September 2018). Najera M, Levison P. (2017). Multicolumn chromatography: facilitating the commercialization of monoclonal antibodies. http://www.bioprocessintl. com/downstream‐processing/chromatography/ multicolumn‐chromatography‐facilitating‐ commercialization‐monoclonal‐antibodies (accessed 15 September 2018). Ötes, O., Flato, H., Vazquez Ramirez, D. et al. (2018). Scale‐up of continuous multicolumn chromatography for the protein a capture step: from bench to clinical manufacturing. J. Biotechnol. 281: 168–174. Brower, M. (2011). Working towards an integrated antibody purification process. IBC Biopharmaceutical Manufacturing and Development Conference, San Diego, CA (September).

27 Makoviecki, J. and Mallory, H. (2013). Adjusting pH

during viral inactivation. GEN 33 (8): 38–39.

28 Gillespie, C. (2018). Condsiderations for in‐line viral

29

30

31

32

33

34

35

36

inactivation. Continuous Biomanufacturing: Achievements and Challenges for Commercial Implementation, Oxford (20–22 June). Pall Life Sciences. (2016). Pegasus prime: robust retention after pressure interruptions (start and stop). http://www.bioprocessintl.com/wp‐content/ uploads/2016/10/Pall_Prime‐Stop‐Start_AN‐1.pdf (accessed 15 September 2018). McAlister, M. (2017). Virus filtration in continuous bioprocessing: considerations for filter design space and validation. BPI, Boston (9–12 September). Bisschops, M. (2017). Regulatory aspects of continuous bioprocessing. Continuous Biomanufacturing: Current Success and Future Trends, Oxford (26–28 June). Ayturk, E. and Marshall, J. (2012). Using technology to overcome bioprocessing complexity. BioProcess Int. 14 (6 Suppl.): 20–24. Warikoo, V., Godawat, R., Brower, K. et al. (2012). Integrated continuous produciton of recombinant therapeutic proteins. Biotechnol. Bioeng. 109 (12): 3018–3029. Godawat, R., Konstantinov, K., Rohani, M., and Warikoo, V. (2015). End‐to‐end fully integrated continuous production of recombinant monoclonal antibodies. J. Biotechnol. 213: 13–19. Klutz, S., Lobedann, M., Schwan, P. et al. (2015). Developing the biofacility of the future based on continuous processing and single‐use technology. J. Biotechnol. 213: 120–130. Schofield, M., Rogler, K., Gjoka, X., and Ayturk, E. (2014). Productivity and exonomic advantages of coupling single‐pass tangential flow filtration to multi‐ column chromatography for continuous processing. https://shop.pall.com/INTERSHOP/web/WFS/PALL‐ PALLUS‐Site/en_US/‐/USD/ViewProductAttachment‐ OpenFile?LocaleId=&DirectoryPath=pdfs%2FBiopharm aceuticals&FileName=14.9528_USD3002_Coupling_ SPTFF_Chrom_AN.pdf&UnitName=PALL. (accessed 22 February 2019).

127

10 Application of Microporous Filtration in Single‐Use Systems Christian Julien and Chuck Capron Meissner Filtration Products, Camarillo, CA, USA

10.1 ­Introduction The field of microporous filtration, encompassing the  separation of suspended particles such as solids and  microorganisms from their pharmaceutical fluid streams, deserves careful consideration in the design of suitable and robust filtration solutions. This considera­ tion applies equally to filter solutions deployed as single‐ use assemblies including filters in capsule configurations, as to those used in more conventional multiuse applica­ tions that rely on disposable cartridges installed in fixed stainless‐steel housings. While the fundamental guiding principles of microporous filtration technology remain intact in single‐use product formats, the implementation of best practices differs from conventional reusable applications. This chapter will address the s­uccessful application of the fundamental principles of filtration in single‐use assemblies. Microporous filters are conventionally categorized as membrane filters or depth filters. Membrane filters are composed of a thin semipermeable plastic film referred to as a membrane, which retains particles primarily through surface capture. Depth filters, on the other hand, incorporate porous filtration media which retain particles throughout the media. Within the context of a given filtration application, the performance of a microporous filter is determined by its pore structure and the nature of its material – often polymeric – used to generate the filter medium. One of the critical appli­ cations of microporous filters is the sterilization of large‐volume and small‐volume parenteral solutions, particularly in cases where moist heat sterilization methods are not an option due to the lack of the drug’s thermal stability. Such thermal instability applies to almost all biologics.

Commercial microporous filter configurations fall into two basic form factors depending on the operation, i.e. normal flow filtration (NFF), also known as direct flow or dead‐end filtration, or tangential flow filtration (TFF), also referred to as cross flow. In NFF, the entire upstream feed stream traverses through the filter membrane, retaining particles in the course of the filtration process, and releasing the filtrate downstream of the filter medium. In TFF, the majority of the upstream feed stream passes tangentially across the filter membrane, referred to as the retentate (concentrate), and only a small portion of the feed stream travels through the filter membrane, referred to as the permeate (filtrate). By design, TFF lends itself to time extended or even con­ tinuous processing by avoiding premature fouling of the filter membrane as the surface is swept clear of debris. In comparison, NFF is a batch process which ends when the majority of the capacity of the filter has been utilized. Until a few years ago, single‐use TFF gamma‐irradiated cassettes delivered as preassembled single‐use fluid paths, which eliminate the sanitization and subsequent water for injection (WFI) flush steps commonly required in conventional multiuse systems, was not an established unit‐processing operation. Regardless of whether such devices are configured as plate and frame or as hollow fiber systems, their design, operation, and constituent disposable components such as pumps, fluid paths, sen­ sors, valves, and filter modules are beyond the scope of this chapter. Membrane adsorbers are yet a different cat­ egory of advanced filtration devices and constitute chemically functionalized macroporous membranes that are designed to effect chromatographic separation. Consequently, they are categorized as purification devices under process chromatography and are not fur­ ther discussed in this chapter.

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

128

10  Application of Microporous Filtration in Single‐Use Systems

10.2 ­Microporous Filters 10.2.1  Nominal Versus Absolute Removal Ratings The pore size, shape, or geometry of a microporous filter is determined by the structural voids of its filter media. The pores themselves can be described as a network of irregularly shaped but roughly spherical voids intercon­ nected to form a torturous passageway throughout the structure of the filter. The irregularities in the interpore shape add further complexity to the fluid path. The restrictive pore diameter typically lies within the pore and not at its surface entrance. Microporous filters are often represented with either nominal or absolute removal ratings expressed in microm­ eters. A nominal pore size‐rated filter refers to a filter capable of removing usually between 90 and 99% of solid particles of greater than the stated pore size rating. An absolute pore size rating refers to a filter capable of remov­ ing 100% of a well‐characterized biological particle such as Brevundimonas diminuta for 0.2 μm membranes. While microporous filters are often represented with either nominal or absolute pore size ratings, the actual ratings are seldom derived from direct pore size measurements. Rather, as described above, the pore sizes are inferred by experimentation by challenging the filter with particles or microorganisms of known size and determining their per­ centage retention. A scientifically correct absolute rating would refer to the diameter of the largest spherical particle that will pass through the filter under specified laboratory conditions. Some membranes often used for analytical purposes, typ­ ically referred to as track‐etched membranes because of the manufacturing process, have uniform holes forming cylindrical tunnels through the cross section of the mem­ brane and have a porosity of approximately 10% in a thin sheet of polyester or polycarbonate. Since all particles are captured on the surface, they are ideal for microscopic examination. The absolute rating can be measured with a scanning electron microscope (SEM). In contrast, mem­ branes used for process filtration purposes consist of a three‐dimensional network of interstitial pores that are irregularly shaped. Consequently, their pore size cannot be measured with an SEM. More importantly, the three‐ dimensional network of pores, often combined with natu­ ral or imparted binding characteristics, allow particles smaller than the physical pore to be consistently removed. For these reasons, the particle retention ratings assigned by filter manufacturers depend on the filter type, the material composition of the filter being rated, the type of test contaminant used, the challenge level or the con­ centration of the test contaminant, the differential pres­ sure applied across the filter, and the viscosity of the test fluid. Due to the varying practices among manufacturers

in regard to particle retention test methods, comparison among manufacturers should not be relied upon solely on the basis of the assigned pore size rating. 10.2.2  Particle Retention Mechanisms In practice, microporous filtration solutions deployed in pharmaceutical manufacturing processes are mainly conceived on the basis of particle entrapment through sieve retention. Adsorption sequestration as a comple­ mentary mechanism of retention depends on a variety of factors, but is typically not the predominant particle retention mechanism unless specifically desired. The sieving mechanism can be described based on the size of the particles to be retained relative to the size of the filter pores as shown in Figure 10.1. 10.2.3  Filter Media At the core of a microporous filter, such as in a capsule device with a pleated membrane geometry depicted in  Figure  10.2, is a microporous membrane commonly manufactured from a polymeric material. Common materials used in microporous filters are membranes of mixed cellulose esters (MCE), often a mixture of ­cellulose acetate (CA) and cellulose nitrate (CN), CA, polyether­ sulfone (PES), polytetrafluoroethylene (PTFE), polyvi­ nylidene difluoride (PVDF), and polyamide (Nylon); or depth filter material, such as a nonwoven mat commonly made out of glass fibers (GF), or cellulose (CE), polypro­ pylene (PP), polyethylene (PE), polybutylene terephtha­ late (PBT), or polyethylene terephthalate (PET) fibers. Not all aforementioned filter media are gamma stable or available in gamma‐stable product configurations. For example MCE, CA, and PTFE membrane filters, as well as PP depth and membrane filters, are not commonly used in single‐use assemblies. CE, PTFE, and PP are among those materials that degrade by gamma irradia­ tion albeit to various degrees; CE becomes less durable above 30 kGy, and PTFE becomes significantly damaged when irradiated even at low dosages [1]. Materials of construction that are suitable for microporous filter media intended for single‐use implementations, with the exception of PTFE  –  incorporated here because of its prevalence in industry in autoclavable assemblies – are listed in Table  10.1. Microporous filter media can be organized in regard to their inherent material properties and their manufacturing process, ­enabling one to discern the relative pros and cons of the different microporous filter media in a particular application. The knowledge base created from a material design space is an essential prerequisite in the determination of suitable filter solutions in pharmaceutical applications. Filter media vary in their type, morphology, and pore

10.2  Microporous Filters

Figure 10.1  Description of the sieving mechanism under consideration of size of particles and filter pores.

Sieving mechanism types: • Cake formation: When suspended particles indigenous in the process fluid are larger than the filter’s surface pores, sieve retention will result in the accumulation of suspended particles on the filter’s surface, i.e. cake formation. The permeability of the filter cake depends on the particle packing, the particle concentration, quantity, size, shapes, and deformability under applied differential pressure, to mention the most important factors. The development of a filter cake does not have to be looked upon as an undesirable characteristic because it can be beneficial, allowing the progressive retention of smaller particles, and essentially functions as a prefiltration step, though this may be at the expense of achieving favorable flow characteristics. • Catastrophic pore blockage: When suspended particles indigenous in the process fluid approach the size of the filter’s pores, a phenomenon called catastrophic pore blockage can result due to the close fit between the particle size and the pore size. This situation exposes the retained particles to the full shear force resulting from the applied differential pressure and in cases where deformability of such particles is possible, particle breakthrough may occur. • Adsorption sequestration: Finally, when suspended particles indigenous in the process fluid are smaller than the filter’s pores, the particle may still become captured by contact with the pore walls in a phenomenon known as adsorptive sequestration. Adsorption sequestration contributes to the overall particle retention only when the smallest particle in its distribution is smaller than the biggest pore in the filter pore size distribution. Intrapore adsorptive sequestration may lead to subsequent smaller particle capture by sieve retention as a result of the constriction of the filter’s pores.

Figure 10.2  Cutaway view of a microporous capsule filter with pleated membrane geometry. Fluid enters from the bottom of the capsule, flows around the outside of the pleat pack module, traverses through the filter membrane to the inner core, and exits at the top. Source: Courtesy of Meissner Filtration Products, Camarillo, California.

Vent/drain

Polypropylene housing

Polypropylene support cage

Polypropylene core

Pleated filter membrane, support and drainage layers

129

Table 10.1  Commonly used microporous filter media in single‐use filtration applications organized per filter type featuring information on their filter media manufacturing process, morphology and pore structure formed, and pros and cons. Filter Filter medium medium manufacturing process

Morphology and pore structure formed

PES

Solution chemistry, phase inversion

Asymmetric with broad High flow rate and throughput pore size distribution Low extractables Moderate‐to‐high chemical resistance Wide pH range 1–14 Capable of producing sterile filtrates Capable of retaining mycoplasma Capable of retaining viruses Gamma stable

Moderate‐to‐low nonspecific adsorption depending on surface modifications

PVDF

Solution chemistry, phase inversion Hydrophobic membranes undergo surface treatment (grafting) to render them hydrophilic

Symmetric with narrow Low nonspecific absorption pore size distribution Low extractables Moderate‐to‐high chemical resistance pH range 1–12 Capable of producing sterile filtrates Capable of retaining mycoplasma Capable of retaining viruses Gamma stable

Moderate flow rate and throughput

Membrane PTFE Hydrophobic

Extruded films by a stretching process

Symmetric with narrow High flow rate and throughput High nonspecific adsorption due to pore size distribution Low extractables hydrophobic interactions High chemical resistance Difficult to wet with water Not gamma stable

PVDF

Solution chemistry, phase inversion Varying thinness of membrane

Symmetric with narrow Low nonspecific absorption pore size distribution Low extractables Moderate‐to‐high chemical resistance pH range 1–12 Capable of producing sterile filtrates Capable of retaining mycoplasma Capable of retaining viruses Gamma stable

PP

Spunbonded fibers, calendered, layered, and point bonded

Graded density by varying fiber packaging Typical fiber size 20–40 μm

Filter type

Membrane Hydrophilic

Depth filter

Melt‐blown microfibers, Graded density by varying fiber diameter calendered, layered, Typical fiber size and point bonded 1–3 μm

Pros

High dirt‐holding capacity High flow rate and throughput Low extractables High chemical resistance

Cons

Moderate flow rate and throughput

High nonspecific absorption due to hydrophobic interactions Generally not gamma stable below pore size ratings of 10–40 μm

PBT

Melt‐blown microfibers, Graded density by calendered and layered varying fiber packaging Typical fiber size 1–3 μm

Moderate resistance to High dirt‐holding capacity High flow rate and throughput acids and bases Moderate‐to‐low nonspecific adsorption Gamma stable

Glass fibers (GF)

Resin‐bonded microfibers

Low‐to‐moderate nonspecific Fiber migration adsorption High dirt‐holding capacity High flow rate and throughput High chemical resistance Gamma stable

Typical fiber size 8–16 μm

The PTFE filter medium is included here due to its prevalent use in industry though disposable filter capsules cannot be sterilized through gamma irradiation, rather they must be autoclaved.

10.2  Microporous Filters

Table 10.2  Common terms used to describe filter design and performance attributes. Term

Description

Units

Differential pressure

The difference in pressure between the upstream (feed or influent) and downstream (effluent) sides of the filter. (Synonym: delta P (ΔP), psid or pressure drop)

psi bar, mbar

Dirt‐holding capacity

The quantity of contaminant a filter can remove and hold before the maximum allowable pressure differential or delta P level is reached. (Synonym: contaminant capacity)

g, kg oz.

Effective filtration area (EFA)

The total surface area of the filter available to process fluid

cm2, m2 ft2

Flow rate

The volumetric rate of flow of a solution. The filter manufacturer typically provides flow rate data for a filter at a specified differential pressure. Hydrophilic membrane filters are usually tested with water. Hydrophobic membranes used for air filtration are usually tested with dry air.

l/min, ml/min m3/min, m3/h ft3/min (cfm) gal/min (gpm)

Flux

Unit of flow per unit of area (Synonym: permeability or flow density)

l/min/m2, ml/min/cm2

Pore size distribution

Ratio of the number of pores (holes) of a given size to the total number of pores per unit of area

% as function of pore size

Porosity

Ratio of pore volume to total volume of filter media. (Synonym: void volume or open area of filter)

%

Retention efficiency

A measurement of how well a filter retains particles. It is usually expressed as the retention of particles of a specific size by a filter. (Synonym: removal efficiency)

%

Throughput capacity

The amount of solution that passes through a filter. The filter manufacturer typically provides the expected amount of liquid, usually water that a user should be able to pass through the filter prior to the filter plugging.

l gal

Asymmetry ratio

Ratio of the upstream pore size to that of the downstream pore size. (Synonym: gradient density)

Source: Data from Parenteral Drug Association [5].

size ratings, as well as in their chemical compatibility and degree of nonspecific adsorption with the feed stream. Common terms used to describe filter design and per­ formance attributes are listed in Table 10.2. Selecting a filter to achieve the desired performance outcome, for example, in terms of flow rate or throughput, in process­ ing a given feed stream, is directly related to the filter media and their manufacturing process. 10.2.4  Membrane Filters The pores are the result of the inter spacing of molecular chains within polymeric materials and their physical and chemical processing thereof. Microporous membranes are typically created by dissolving a polymer in a solvent and then reprecipitating the polymer back into a solid by exposure to a non‐solvent. The process is carefully con­ trolled to create the desired pore structure and great care is taken in well‐crafted membranes to make the pore size and geometry as consistent and uniform as possible. The  structure of membrane‐based microporous filters is  either symmetric, also referred to as isotropic, or

a­ symmetric, also referred to as anisotropic. Symmetric membranes have a uniform pore size structure through the entire membrane, and they can be more or less porous. The resistance to mass flow in symmetric mem­ branes is determined by the pore size, the porosity, and the total thickness of the membrane, i.e. a decrease in the membrane thickness results in an increase in flow rate. Asymmetric membranes are characterized by a gradient of pores sizes through the membrane, meaning that the pores start larger near one surface of the membrane and then become gradually smaller. Asymmetric membranes are capable of higher flux and higher capacity than ­isotropic membranes in some applications. The structure of a PTFE membrane shows intermittent long separated strands of stretched polymer film such as that depicted in Figure  10.3. The pore size of a PTFE membrane is the direct result of the amount of stretch that has been applied to the polymeric film that lay at its origin. Manufacturers of PVDF and PES membranes have each developed proprietary formulas of casting solutions and unique manufacturing processes that yield reproduc­ ible membrane structures of consistent pore size. Because

131

132

10  Application of Microporous Filtration in Single‐Use Systems

Figure 10.3  Microporous PTFE filter, 0.2 μm rated at 2000× Source: Courtesy of Meissner Technical Services, Meissner Filtration Products, Camarillo, California.

Figure 10.4  Microporous isotropic PVDF membrane, 0.2 μm rated at 3500× Source: Courtesy of Meissner Technical Services, Meissner Filtration Products, Camarillo, California.

the solution chemistry and the manufacturing process determine the resulting membrane characteristics, there are no two membranes from different membrane manu­ facturers that are exactly alike. Figure 10.4 shows a sym­ metric PVDF membrane and Figure 10.5 shows a highly asymmetric PES membrane. As can be readily derived from the SEM images, the pore size shape and geometry make it impossible to measure the pores sizes directly. Membrane filters are characterized by the largest pore in symmetric membranes or in the retentive layer in asymmetric membranes, along with the mean flow pore according to American Society for Testing and Materials (ASTM) Standard F‐316 [2]. The actual pore size distri­ bution in membrane filters remains technically chal­ lenging to determine and is usually unknown, though it is the total porosity of the membrane as well as the membrane’s inherent hydrophilicity that determines the flow characteristics. Filters can also be composed of stacked membranes, each layer characterized as having the same or different

Figure 10.5  Microporous highly asymmetric PES membrane, 0.2 μm rated at 1000× Source: Courtesy of Meissner Technical Services, Meissner Filtration Products, Camarillo, California.

pore sizes. In the case of progressively stacking filter membranes with the same pore size ratings, the reten­ tion efficiency is squared, but the absolute rating remains unchanged. For example, if a membrane layer is 90% effi­ cient, the composite rating of two layers will be 99% but the absolute rating remains unchanged. The successive layering will ultimately result into a plateau value in regard to the sieve retention of the composite. Creating stacked membranes with different pore size ratings and stacking them from larger to narrower pore size ratings relative to the direction of the feed stream is intended to affect a degree of asymmetry for prefiltration. Such stacked membranes often increase the contaminate holding capacity by allowing larger particles to be retained in larger pores preserving small pores for smaller contaminants. Two‐ and even three‐layer ­composite membrane filters are typically used in the construction of a stacked membrane. 10.2.5  Depth Filters The pores of depth filters are the result of a torturous path or interstices among the matrix created by the fibers used. Depth filters are often characterized in terms of their dirt‐holding capacity which is the direct result of their manufacturing process being able to instill a more or less graded fiber density. A graded fiber density can be achieved in a spunbond process by varying the fiber den­ sity or in a melt‐blown process capable of incorporating composite layers of media as described above and calen­ dering them to produce a single asymmetric structure. Because the fibers are randomly deposited in the manu­ facturing process, the resulting pore size distribution is significantly broader as compared to a cast solution pro­ cess used to manufacture membrane filters. This broader pore size distribution is the main reason that depth filters cannot be relied upon to produce a sterile filtrate.

10.2  Microporous Filters

Fiber‐based media are characterized by the fiber diam­ eter, length, density, and the bonding method. The spe­ cific material composition chosen along with the manufacturing process of the fiber‐based media and the applied process controls, particularly surrounding the thickness of the fibers, are crucial for its suitability in various applications. Properties like fiber diameter and brittleness should be carefully assessed in conjunction with the effects of gamma irradiation prior to the deploy­ ment of depth filter media in single‐use applications. Fiber generation may be compounded by the radiolytic degradation effects of gamma sterilization. In general terms, depth filters should be considered as potentially fiber releasing because they are composed of fiber layers, a property that cannot necessarily be elimi­ nated by pre‐flushing the filter with liquid. To reduce or eliminate this risk, fibers are generally continuous and should be non‐brittle to prevent downstream migration. In addition, final membrane filters may be incorporated downstream of depth filters to mitigate the risk of fiber migration in pharmaceutical drug manufacturing. High‐capacity depth filtration is used for the clarifica­ tion of cell culture harvests where the objective is to sepa­ rate cells or their lysates to produce a clear supernatant. Lenticular filter designs that constitute fully enclosed stacked disks, with mechanical compression seals and interlocking modules, in sizes up to several m2 in EFA have become popular. Filter media are typically based on resin‐bonded CE fibers and the use of filter aids, such as diatomaceous earth, is common. Although often classified as disposable devices, these filters require pre‐use flushing and must be sanitized or autoclaved because they cannot be sterilized by gamma irradiation. The term disposability is used here in a more colloquial meaning and may have more to do with operator convenience and safety, than implementing a fully single‐use operating paradigm. 10.2.6  Sterilizing‐Grade Filters When particle removal entails the complete removal of microorganisms from a fluid, the filtration process is being referred to as sterilization. Sterilizing‐grade filters often, as the case may be, carry a 0.2 or 0.22 μm pore size rating, and are considered interchangeable [3]. However, caution is warranted because sterilization is not an inher­ ent property of any particular pore size designation. Because 0.2 μm‐rated filters may differ, among other properties, in their retention, the Food and Drug Administration (FDA) issued guidance that steers clear of actual pore size designations and defines a sterilizing‐ grade filter based on a critical performance characteris­ tic, i.e. a sterilizing‐grade filter is defined as one that demonstrates complete removal of 107 B. diminuta (ATCC 19146) per square centimeter of EFA under

worst‐case conditions – those that are least conducive to organism retention [3]. Consequently, filter manufactur­ ers subject their labeled sterilizing‐grade filters to a microbial challenge of at least 107 per square centimeter B. diminuta (ATCC 19146) and test following ASTM F838 methodology [4, 5]. Live organisms are chosen because of the ability to challenge the membrane with billions of “particles,” capture any that pass through the membrane, and culture them so that if even one should pass it will be detected. Testing with laser particle coun­ ters is a statistical process which does not lend itself to extremely high challenge levels capable of detecting that every particle has been removed. Notwithstanding a supplier’s demonstration of sterilizing‐grade filter capa­ bility, the reliability of a sterilizing grade filter to sterilize a given process stream under actual‐use conditions must be validated under worst‐case process conditions and is a regulatory requirement [3]. This requirement stems from the knowledge that a supplier‐qualified sterilizing‐ grade filter may not always produce a sterile effluent under all conditions of use and in all process streams. To avoid grow‐through, a phenomenon where micro­ organisms retained on the upstream side of a micropo­ rous sterilizing‐grade membrane penetrate it after a period of time, resulting in the passage of organisms downstream, the usage time of the filter is generally lim­ ited to approximately eight hours. This time frame is considered a safe practice and also falls within typical durations of pharmaceutical filtration applications. Longer filtration times are possible but must be vali­ dated. A grow‐through occurrence may be the result of a morphological reduction in size that certain organisms undergo after exposure to particular drugs. Hereto, the  physicochemical properties of the process fluid serve  to alter the size of suspended organisms. Grow‐ through  –  although in practice avoided by the time‐ restriction set on processing time – remains an ongoing matter of concern, largely because its mechanism is still little understood and its control, therefore, is uncertain. For example, the size of organisms could differ depend­ ing upon the stage of their cellular development and upon how they were cultured. Nevertheless, a growth‐ through occurrence can be avoided by undertaking expe­ ditious filtrations, setting time restraints, utilizing larger EFA’s, or more frequent filter change‐outs. Desired filter retention should thus be ensured by validation, regard­ less of the possibility of grow‐through [6]. Likewise, other process activities such as prolonged filling times will receive their needed documented experimental veri­ fication. Regardless, the possibility of grow‐through will not pose a threat in pharmaceutical filtrations, due to the requirement for validation. Whenever a filter is used to sterilize a fluid, validation is required to demonstrate that it will perform satisfactorily, reliably, and repeatedly

133

134

10  Application of Microporous Filtration in Single‐Use Systems

over the entire processing interval even when exposed to worst‐case processing conditions. Thereby, grow‐ through, if it were to occur, will be detected. 10.2.7  Mycoplasma Retentive Filters Because a sterilizing‐grade filter is modeled in its reten­ tion using B. diminuta, it may not retain yet smaller organisms that may be encountered in pharmaceutical process streams. Mycoplasma is one such organism. Mycoplasma is a genus of bacteria that lacks a cell wall and is known to spread quickly in cell cultures. Not sur­ prisingly, mycoplasma is often associated with media, sera, and reagents. In such cases, smaller pore size ratings such as 0.1 μm – or even 0.04 μm – can be used, such as is modeled by the retention of Acholeplasma laidlawii [7]. 10.2.8  Virus Retentive Filters Viral safety in biopharmaceutical manufacturing is a pre­ dominant consideration and a regulatory requirement [6, 8]. Among the various orthogonal techniques available today to risk mitigate the presence of endogenous viruses in source materials and adventitious viruses in the pro­ cess, viral clearance through virus retentive filters has become a center‐stage unit operation [9]. These filters are  also often referred to as nanofiltration devices and typically fall into two categories. The first category ­ encompasses filters capable of log reduction value (LRV) of minimum 3–6 for large viruses, which are typically 80–110 nm endogenous retroviruses. The second ­category consists of filters capable of a LRV of a minimum of 3 for small parvoviruses, which are typically 18–24 nm – these filters will also retain larger viruses. This second category of filters requires a very narrow pore size distribution to ensure high recoveries of the protein, which in the case of monoclonal antibodies – among the largest therapeutics biologics  –  have hydrodynamic radii in the range of 5–7 nm [10]. Virus retentive filters are commonly com­ posed of either NFF or TFF configurations with regener­ ated CE, PVDF, or PES filter media in sizes up to 4 m2 per device. With the exception of some newer PVDF‐ and PES‐based filters that are compatible with gamma irradia­ tion, lending them to be readily incorporated in single‐use assemblies, most filters are sanitized or sterilized by auto­ claving to reduce bioburden.

10.3 ­Filter Selection 10.3.1  The Need for Filter Testing The current state in filter technology is not able to fully characterize membrane and depth filters in terms of their pore size distribution. In addition, all too often, little is

known about the particles in the fluid stream to be f­ iltered, further rendering a suitability assessment for a filter on the premise of assessing both pore size and particle size distributions, while immensely desirable, in practice, not attainable. Because the interaction of a given membrane and depth filter to a particular fluid stream under specific process conditions cannot be fully predicted, filter testing becomes the practical means of selecting membranes and depth filters suitable for their intended purpose. Factors that can affect filter performance include the viscosity, surface tension, osmolality, ionic strength, and pH of the fluid stream to be filtered, the chemical interaction of the fluid stream with the filter, and the operational parame­ ters including pressures, flow rates, maximum use time, and temperature [3]. The particles encountered in phar­ maceutical fluid streams are typically irregularly shaped and are often deformable, representing quite a departure from the idealized models which use rigid spherical parti­ cles. Therefore, in practice, many possibilities may exist to find a suitable filter solution given a specific application. In this context, a narrow or broad pore size distribution is inherently neither good nor bad, though its successful deployment depends entirely on the filter problem that needs to be solved. Suspended particles in pharmaceutical fluid streams do not typically represent themselves as mono disper­ sions; rather they are characterized by a wider particle distribution that is often unknown. Therefore, a com­ mon practice is to deploy multiple filters in a combina­ tion from larger to smaller pore size ratings to achieve higher throughput capacity. The upstream filter is known as the prefilter. Prefilters can be microporous depth or membrane filters and are selected for their ability to remove particles that would tend to plug the downstream filter. This series of filters is known as a filter train. A well‐balanced filter train, as are those derived from filterability trials, represents an economical process ­ solution because it extends the life of the more‐expen­ sive membrane filters and optimizes the filter areas of each of the constituting filters. Typically, prefilters with larger pore sizes are more cost‐effective and thereby can economically contain larger EFAs as compared to ­ ­membrane filters with smaller pore sizes. 10.3.2  Flow Decay Studies Filter suitability assessment and sizing is based on data derived from flow decay measurements as a function of time. The goal is to determine the maximum volume that can be filtered through a given filter. Because liquid flow through microporous filter media is proportional to the applied differential pressure across the membrane and the EFA, and is inversely proportional to the liquid vis­ cosity and the filter membranes’ resistance – as described by Darcy’s law – the flow rate will decline over time as a

10.3  Filter Selection

result of the increased membrane resistance due to grad­ ual pore plugging. This phenomenon holds true for feed streams that are high in particles where filter capacity is the limiting factor. Flow decay studies  –  also known as flux decline or simply filterability studies  –  are most commonly per­ formed using the constant pressure filtration method, typically utilizing differential pressures between 5 and 30 psid, with the results being conveyed by plotting the filtrate volume or throughput as a function of time. The declining flux during filtration is specific to the filter/ filtrate combination tested and will differ according to  the particle removal mechanism whether it is due to  sieve retention with cake formation, or adsorption sequestration with consequent pore narrowing and blocking. Alternatively, although less commonly used, filterability studies can be performed using the constant volume filtration method, whereby the pressure of the liquid to be filtered is varied to maintain a constant flow rate of the filtrate. Flow decay studies are conducted on small‐scale filter devices that are representative scaled‐ down models of the larger filter capsules. Smaller filter devices are often preferred because they allow for shorter test times. Commonly, to model flow decay in mem­ brane filters, 47 mm or smaller membrane disc filters are used, or in the case of depth filters, small capsule filters are used. The advantage of using scaled‐down devices is to limit the amount of liquid needed while preserving the predictability of these devices for subsequent scale‐ up calculations. 10.3.3  Meeting Process Objectives From a pharmaceutical production perspective, micro­ porous filters are purposely selected to meet specific process objectives. Common objectives for a biological process include the removal of particle and microbial contaminates from raw material feed streams, such as cell culture media, sera, and buffers; the primary and secondary clarification of bioreactor harvest streams; the bioburden reduction/control of intra process trans­ fers of value‐added feeds, such as intermediates across multiple‐unit process operations; viral clearance; and the sterilization of final drug product. In addition, filtration solutions may be sought after that do not remove valua­ ble products through nonspecific absorption, and those that do not introduce new contaminates to the fluid streams like filter extractables. The latter issue is often addressed by pre‐flushing the filter. One of the main objectives of a microporous filtration scheme is to remove particles from a feed stream within a specified process time to meet the requirements set out by the overall pharmaceutical production step, with the largest possible volumetric throughput, in the most eco­ nomical way possible. Herein lays the importance to

(a)

(b)

(c)

Figure 10.6  Single‐use filter device configurations from left to right: inline capsule filter (a); T‐style capsule filter (b); and modular configuration composed of multiple T‐style capsules filters (c). Source: Courtesy of Meissner Filtration Products, Camarillo, California.

conduct material compatibility and filterability studies that provide the necessary data to establish process feasi­ bility and enable scale‐up calculations for filter sizing that meet the performance requirements. Only in cases where filter capacity is not the limiting factor, such as in feed streams that are low in particles, such as, for exam­ ple, typically in process buffers, water, and gases, can ­filter sizing be based on manufacturer’s provided flow rate data. Once the filter media and sizing have been selected for a given application, end users can often choose the device configuration, i.e. inline versus T‐style capsule filters or completely integrated modular devices configurable in various capsule filter combinations as shown in Figure 10.6 to suit their desired form factor. A particular process objective for hydrophobic vent fil­ ters is to avoid or eliminate condensate buildup which can cause blockage of the membrane and subsequently lead to higher back‐pressures. Back‐pressure requires thoughtful consideration in bioreactors and must be carefully monitored and controlled. This holds especially true in single‐use bioreactors due to the limited hoop strength of the bags utilized, especially with larger form factors. The use of filter capsule heaters, bidirectional flow membranes, parallel filter configurations, and/or liquid foam traps have all been successfully used to miti­ gate back‐pressure buildup during the course of the culture. The main process objective for a virus‐retentive filter process is to achieve reliable and robust clearance of virus which is established in viral clearance studies. An immediate secondary goal is to utilize a process filter that maximizes protein recovery. Because of the signifi­ cant consumable cost associated with viral retentive filters, the throughput is optimized during process ­

135

136

10  Application of Microporous Filtration in Single‐Use Systems

­ evelopment. Hereto, the filter capacity and flux is opti­ d mized for a given protein concentration in the feed stream, which itself depends on the relative position in the downstream process purification scheme. For exam­ ple, a lower protein concentration feed stream requires less EFA but the filter needs to process an increased vol­ ume because the feed stream is more diluted. The impact of protein concentration on the EFA needed can be so profound that multiple flow decay studies for varying protein concentrations are often conducted. In addition to the effect of protein concentration on the EFA of the filter, the impurity levels, protein aggregates, residual deoxyribonucleic acid (DNA), and other trace contami­ nants in the feed stream can have an equally dramatic impact on the filter performance, hence the need for optimized prefiltration such as, for example, the use of adsorptive depth filters. Viral retentive filters are quali­ fied for their viral clearance capability in spiking studies conducted on scaled‐down laboratory models of actual pharmaceutical manufacturing process [6, 9]. The vali­ dation of the scaled‐down model necessitates the availa­ bility of small‐scale virus retentive filters that are representative of their larger production siblings and that can be operated at the same flux and differential pressures and with demonstrated equivalency in terms of their virus LRV, protein recovery, and quality. Parvovirus breakthrough continues be a concern among end users, regulators, and filter manufacturers alike because the mechanism and critical filter process parameters are not fully understood [11]. As a result, whenever a parvovirus retentive filter is chosen as the worst‐case model for viral clearance, it is essential that clearance studies evaluate the potential risk for parvovi­ rus breakthrough. In order to avoid an adverse impact on the filter flux, it is critical to use highly purified virus preparations in the viral clearance studies [12]. 10.3.4  Applications Orientation Notwithstanding the need for filter testing, it is possible to develop a high‐level orientation regarding micropo­ rous filter applications that are commonly encountered by pharmaceutical end users. Such an overview borrows from our knowledge of filter media properties and proven performance in a multitude of pharmaceuticals manu­ facturing processes, although it can only serve as an ini­ tial guide. Table 10.3 lists common filter media, organized per pore size ratings, and corresponding filter applica­ tions. Both microporous membrane and depth filters are commercially available in a range of pore size ratings as single or dual layers, or as multimedia configurations. Multimedia filter configurations often combine different prefilters and final filter media, such as, for example, a GF prefilter with a PVDF final filter (GF/PVDF) or a PES prefilter with a PVDF final filter (PES/PVDF).

10.4 ­Final Sterile Filtration 10.4.1  Regulatory Highlights Although redundant filtration is not specifically stated as a regulatory requirement, the FDA guidance does men­ tion “Use of redundant sterilizing filters should be con­ sidered in many cases” [3]. The European Union (EU) guideline states that “Due to the potential additional risks of the filtration method as compared with other sterilization processes, a second filtration via a further sterilized micro‐organism retaining filter, immediately prior to filling, may be advisable. The final sterile filtra­ tion should be carried out as close as possible to the fill­ ing point” [13]. The European Medicines Agency (EMA) guidance states that “For sterilization by filtration, the maximum acceptable bioburden prior to filtration must be stated in the application. In most situations NMT (Not More Than) 10 CFU/100 ml will be acceptable, depending on the volume to be filtered in relation to the diameter of the filter. If this requirement is not met, it is necessary to use a pre‐filtration through a bacteria‐retaining filter to obtain a sufficiently low bioburden” [14]. It is not a surprise that the NMT 10 CFU/100 ml is being interpreted by many end users as a requirement to add a sterilizing‐grade filter as a prefiltration step, whereas demonstrating process capability and feed bioburden control would have been more commensurate with a modern risk‐based validation approach. In con­ trast, a single‐stage final filter layout, while advantageous in terms of holdup volume, cost, required flush volume, and operational simplicity, does not provide control of the bioburden feed stream. In addition, a risk assess­ ment, in regard to a single‐stage final filtration setup, may point to a higher level of process risk in the case of a filter integrity failure as compared to a serial or redun­ dant filtration scheme. 10.4.2  Serial and Redundant Filtration In a serial filtration setup, the addition of a second steri­ lizing‐grade filter, either upstream of the final filter or the sterile holding bag, imparts the necessary bioburden control. The most common implementation is to use a redundant filtration setup whereby two sterilizing‐grade fitters are used, one immediately upstream of the other, and where each filter is validated to provide a sterile fil­ trate. The terms serial and redundant filtration are often used interchangeably although this can be inaccurate. Serial filtration occurs when two or more filters are used one after the other. The filters may have the same or decreasing size pores. Redundant filtration is a special case of serial filtration where two sterilizing‐grade f­ ilters are used.

10.4  Final Sterile Filtration

137

Table 10.3  Commonly used microporous filter media in single‐use assemblies, organized per filter medium, pore size ratings, and corresponding filter applications. Filter medium Filter type PES

PTFE

PVDF

PVDF

Filter medium attributes

Filter pore size ratings Corresponding filtration applications

Membrane

Virus retentive

20 nm

Viral clearance

Hydrophilic

Mycoplasma retentive High flow rate and throughput

0.04 μm 0.1 μm 0.2/0.1 μm

Filtration of cell culture media and serum

High flow rate and throughput Moderate‐to‐low nonspecific adsorption Low extractables

0.2 μm 0.4/0.2 μm 0.8/0.2 μm

Sterilization of aqueous‐based pharmaceutical solutions Sterilization of cell culture media, buffers, and aqueous process solutions

Wide pH range 1–14

0.2 μm 0.4 μm 0.6 μm 0.8/0.4 μm

Clarification, particle removal, and bioburden reduction Prefiltration of various aqueous‐based solutions

Membrane High flow rate and throughput Hydrophobic

0.1 μm 0.2 μm

Sterilization of gas inlet and vent, single‐use bioreactors, and solvents Autoclavable assemblies only

High flow rate and throughput High nonspecific adsorption Low extractables High chemical resistance

0.4 μm 1.0 μm 3.0 μm 5.0 μm

Clarification, particle removal in organic solvents, acids, and bases Autoclavable assemblies only

0.2 μm

Sterilization of gas inlet and vents, single‐use bioreactors

0.4 μm

Clarification, particle removal in gas

Mycoplasma retentive Moderate flow rate and throughput

0.1 μm 0.2/0.1 μm

Filtration of cell culture media and serum

Moderate‐to‐high flow rate and throughput Low nonspecific adsorption Low extractables Moderate‐to‐high chemical resistance pH range 1–12

0.2 μm 0.4/0.2 μm

Sterilization of pharmaceutical solutions Sterilization of buffers and media Final filtration of bulk drug product, single‐use filling assemblies

0.4 μm 0.6 μm 0.6/0.4 μm 5 μm

Clarification, particle removal in biological solutions Clarification, particle removal in certain organic solvents

Membrane Moderate flow rate Hydrophobic Moderate‐to‐high flow rate and throughput Low nonspecific adsorption Low extractables Moderate‐to‐high chemical resistance pH range 1–10 Membrane Hydrophilic

PP

Depth filter

High dirt‐holding capacity High flow rate and throughput High nonspecific adsorption Low extractables High chemical resistance Wide pH range 1–14

40 μm 70 μm

Clarification, particle removal in biological solutions Clarification, particle removal in organic solvents, acids, and bases

PBT

Depth filter

High dirt‐holding capacity High flow rate and throughput Moderate‐to‐low nonspecific adsorption Moderate resistance to acids and bases

0.6 μm 1.2 μm 2.4 μm 5 μm 7 μm 10 μm 20 μm

Clarification, particle removal in biological solutions

GF

Depth filter

High dirt‐holding capacity High flow rate and throughput Low‐to‐moderate nonspecific adsorption High chemical resistance

0.5 μm 1.0 μm

Clarification, particle removal in biological solutions

Pore size ratings provided are either for single‐layer filters or dual‐layer filters where the first rating corresponds to the prefilter and the second rating to the final filter separated by a forward slash. The PTFE filter medium is included here due to its prevalent use in industry though disposable filter capsules cannot be sterilized through gamma irradiation, rather it must be autoclaved. PP depth filters are not gamma stable below pore size ratings of 10–40 μm, rather they must be autoclaved.

138

10  Application of Microporous Filtration in Single‐Use Systems

Both serial and redundant filtration setups add opera­ tor complexity, cost, and lead to higher holdup volumes. Nevertheless, a benefit of choosing a redundant filter setup, as compared to a serial setup, lies in the ability to recover a batch if one filter fails integrity. True redun­ dant filtration requires careful implementation to ensure no bacterial ingress can occur through the vents or drains located between the two filters. Redundant steri­ lizing‐grade filters can provide an extra margin of safety for final filtration, offering insurance against failure of the primary sterilizing‐grade filter, the last filter in the train. Redundant sterilizing‐grade filters are often con­ sidered for high‐value products as well as those products that cannot be reprocessed. They may be employed when filtration takes longer than eight hours and/or when the first filter in series functions as a prefiltration step, thereby reducing the particle load and increasing the efficiency of the second final filter.

10.5 ­Filter Integrity Testing 10.5.1  Regulatory Highlights Integrity testing is a regulatory requirement for all steri­ lizing‐grade filtration applications. The FDA guidance states “Integrity testing of the filter(s) can be performed prior to processing, and should be routinely performed post‐use. It is important that integrity testing be con­ ducted after filtration to detect any filter leaks or perfo­ rations that might have occurred during the filtration” [3]. The EU guideline states that “The integrity of the sterilized filter should be verified before use and should be confirmed immediately after use by an appropriate method such as bubble point, diffusive flow or pressure hold test” [13]. The two regulatory agencies have differ­ ing views. The FDA mandates post‐use filter integrity testing but not pre‐use filter integrity testing although it does recommend it. Whereas the EU mandates both pre‐ use and post‐use filter integrity testing. The implication of the EU pre‐use filter integrity testing requirement is that integrity testing must be performed without com­ promising the filter’s sterility. 10.5.2 PUPSIT Given the regulatory perspectives, the decision to con­ duct a pre‐use/post sterilization integrity test (PUPSIT) should be risk‐based and may include evaluating factors such as the potential for filter damage during installation and sterilization, the ability to perform the integrity test without compromising the downstream sterility, and the potential for rework of the product if the filter fails integ­ rity post‐use. A risk assessment of a single‐use process

implementation reveals a different outcome than one conducted on a multiuse process implementation using filter cartridges in conjunction with stainless‐steel ­housings. That is evident because single‐use systems, by virtue of their design, are fully closed and gamma irradi­ ation causes less thermal and mechanical stress to a filter as compared to steam sterilization [15]. The variant option to conduct a pre‐use/presterilization test only applies in a multiuse manufacturing implementation because in a single‐use implementation the entire assem­ bly is supplied sterile. In a single‐use process implementation, conducting PUPSIT requires the atmospheric pressure downstream of the filter to be maintained during the test, which is one of the reasons why a diffusive flow (DF) test is preferred over a bubble point (BP) test, while other reasons are the higher sensitivity of a DF test and less effluent buildup. PUPSIT applied in single‐use systems adds design c­ omplexity due to the need to provide fluid paths and componentry for the introduction of the wetting liquid and the sterile test gas upstream of the filter, as well as for the collection of both the filter upstream vent and downstream effluent (gas and liquid). Perhaps, an unintended, but nevertheless added value as a result of performing PUPSIT, is that the filter is flushed by the wetting agent which leads to a reduction in the filter’s extractables profile. The single‐use assembly can further be designed with componentry to accommo­ date a blow down of the filter following PUPSIT to prevent dilution of the product stream by the residual wetting liq­ uid. A filter blow down necessitates exceeding the filter’s BP and adds ­process time. 10.5.3  Filter Integrity Tests There are three types of nondestructive filter integrity testing, i.e. the BP test, the DF test, and the water intru­ sion test for hydrophobic filters. Other tests like the pressure hold, forward flow, and pressure decay tests are simply variations of the diffusion test and are not further discussed. In the BP test, the BP is defined as the applied differen­ tial air pressure across a completely wetted filter mem­ brane where the wetting liquid, usually water or alcohol, is forced out of the largest pores and airflow begins. Increasing the air pressure beyond the BP will cause the next smaller pores to open, until the air pressure is finally sufficient to evacuate the wetting liquid from even the smallest pore. The BP measurement has a direct correla­ tion to the filter’s largest restrictive pores, i.e. the higher the BP pressure value the narrower the pore diameter. The DF test is performed at a pressure about 80% of the BP for structurally isometric membrane filters, and less than 80% for asymmetrically structured membrane filters. When applied under a pressure (below the BP

10.6  Filter Qualification and Validation

pressure) gas from the upstream side of the filter will ­diffuse through the liquid‐wetted membrane. This phe­ nomenon of diffusion across the wetted membrane is the basis for the diffusion test. The rate of this diffusion can be measured, and its value used as an index of the integ­ rity of the filter. DF occurs across the membrane in accordance with Fick’s law of diffusion. The DF does not directly correlate to the size of the pore; rather it does correlate to the thickness of the membrane. The flow of gas, due to even the smallest of imperfections in the membrane that manifest by thinning of the liquid layer, result in increased gas flow. The DF test is more sensitive than the BP test. Filter manufacturers specify the DF in terms of test pressure and a maximum allowable flow value, the validity of which are established by empirical correlation with the bacterial challenge test, performed by the filter manufacturer in accordance with ASTM F838 methodology [4, 5]. After processing pharmaceutical fluids, the traces of product on the filter membrane can be difficult to remove or flush and may result in false integrity test readings. Under these circumstances, users often integrity test their filter in the process fluid. This is known as the prod­ uct‐wet integrity test. Consequently, these product‐ specific integrity test values need to be determined separately by laboratory experimentation, which falls well within the realm of validation services provided by filter manufacturers. Pharmaceutical end users typically perform filter integrity test using automated integrity testers, which are convenient and accurate.

10.6 ­Filter Qualification and Validation 10.6.1  Regulatory Highlights In order to ascertain a filter’s suitability for a pharmaceu­ tical process, filter manufacturers publish qualification data derived from laboratory testing. Such data, among others, may include a filter’s materials of construction, EFA, maximum allowable operational temperature and differential pressure, sterilization methods, chemical compatibility, compendial biosafety, particulate matter, non‐fiber release, oxidizable substances, total organic carbon, conductivity, nonvolatile residue, integrity test, and flow decay data. In addition, for sterilizing‐grade fil­ ters documentation provided includes bacterial chal­ lenge test data according to ASTM F838 methodology and correlation to nondestructive filter integrity tests. The qualification documentation represents a reposi­ tory of useful data and supports the subsequent process validation conducted by end users, but is not intended to replace it. The Parental Drug Association (PDA) Technical Reports TR 26 and TR 40 provide a

c­ omprehensive overview of qualification and validation requirements among filter manufacturers and end users along with test methods and best practices [5, 16]. Following a risk‐based assessment, filter manufacturers may also provide more detailed extractables data derived from model solvent studies for process filters, including sterilizing‐grade filters [17, 18]. The premise is that ­process equipment‐related leachables (PERLs) such as those originating from single‐use systems have the potential to alter a key quality attribute of the drug sub­ stance or drug product should the PERLs persist through the manufacturing process. Such extractables data are going to very helpful in the assessment of leachables. Due to their technical nature filter process, validation studies are often conducted by filter manufacturers in specialized laboratories; however, from a regulatory point of view the responsibility for the review of the filter validation data and acceptance resides with the end user. Filter validation is a regulatory requirement. The FDA guidance states “A sterilizing grade filter should be vali­ dated to reproducibly remove viable microorganisms from the process stream, producing a sterile effluent. Whatever filter or combination of filters is used, valida­ tion should include microbiological challenges to simu­ late worst‐case production conditions for the material to be filtered and integrity tests results of the filters used for the study. Direct inoculation into the drug formulation is the preferred method because it provides an assessment of the effect of drug product on the filter matrix and the challenge organism” [3]. The EU guideline states that “All sterilization procedures should be validated” [13, 14]. This applies to all sterilizing‐grade filters used in the manufacturing of the finished dosage form. An EU draft guideline, undergoing current review, states relative to sterile filtration “The filter should be validated with regards to bacterial retention capacity, solution compat­ ibility and leachable filter materials” [19]. 10.6.2  Product‐Based Tests Product‐based testing includes as a minimum product‐ based integrity testing, and for sterilizing‐grade filters bacterial retention following ASTM F838 methodology, with a microbial challenge of at least 107 B. diminuta (ATCC 19146) per square centimeter of EFA under worst‐ case conditions [4, 5]. Depending on the drug formula­ tion, prior laboratory tests may be needed to ensure that the challenge organism is not impacted by certain drug‐ formulation components that could have a bactericidal activity. In cases where bacterial counts are impacted, the drug formulation can be tested without the interfering component. Additional tests performed as part of a filter process validation include particulate testing of the assembly under actual process conditions, product‐based

139

140

10  Application of Microporous Filtration in Single‐Use Systems

chemical compatibility, leachable, adsorption, and prod­ uct recovery testing.

10.7 ­Summary and Conclusions The application of microporous filtration in single‐use ­systems is anchored in the fundamental principles of microporous filtration technology and the vast reposi­ tory of data available in the scientific literature. These foundations equally apply to the removal of particle and microbial contaminants from raw material feed streams, primary and secondary clarification, bioburden reduc­ tion/control, viral clearance, and the sterilization of final drug product. However, there exist some differences in implementation practices. This stems foremost from the

gamma stability requirement often narrowing filter media to popular choices like PES and PVDF mem­ branes. Fiber‐based depth filters require careful assess­ ment because fiber generation may be compounded by the radiolytic degradation effects of gamma sterilization. Hydrophobic vent filter selection in single‐use bioreac­ tors also requires special attention due to limited back‐ pressure design requirements for bags. In a single‐use process implementation, conducting PUPSITs adds design complexity due to the need to pro­ vide fluid paths and componentry for the introduction of the wetting liquid and the sterile test gas upstream of the filter, as well as for the collection of both the filter upstream vent and downstream effluent (gas and liquid). Given the prevailing regulatory perspectives, the ­decision to conduct a PUPSIT should be risk‐based.

Nomenclature ASTM American Society for Testing and Materials BP Bubble point CA Cellulose acetate CE Cellulose CN Cellulose nitrate DF Diffusive flow DNA Deoxyribonucleic acid EFA Effective filtration area EMA European Medicines Agency EU European Union FDA Food and Drug Administration GF Glass fibers LRV Log reduction value MCE Mixed cellulose esters NFF Normal flow filtration

NMT Not more than PBT Polybutylene terephthalate PDA Parental Drug Association PE Polyethylene PERLs Process equipment‐related leachables PES Polyethersulfone PET Polyethylene terephthalate PP Polypropylene PTFE Polytetrafluorethylene PUPSIT Pre‐use/post sterilization integrity test PVDF Polyvinylidene difluoride SEM Scanning electron microscopy TFF Tangential flow filtration WFI Water for injection

­References 1 Association for the Advancement of Medical

Instrumentation. (2008). AAMI TIR17: compatibility of materials subject to sterilization. Technical Information Report. Arlington, Virginia. International Society of Testing and Materials. (2011). 2 ASTM F316‐03: Standard Test Methods for Pore Size Characteristics of Membrane Filters by Bubble Point and Mean Flow Pore Test. West Conshohocken, Pennsylvania. https://global.ihs.com/doc_detail. cfm?document_name=ASTM%20F316&item_s_ key=00020698 (accessed 22 February 2019). 3 Center for Drugs and Biologics, and Office of Regulatory Affairs, Food and Drug Administration. (1987, revised 2002 and 2004). Guidance for Industry, Sterile Drug Products Produced by Aseptic

Processing, Current Good Manufacturing Practice. Rockville, Maryland. https://www.fda.gov/downloads/ Drugs/Guidances/ucm070342.pdf (accessed 22 February 2019). International Society of Testing and Materials. (2015). 4 ASTM F838‐15a: Standard Test Method for Determining Bacterial Retention of Membrane Filters Utilized for Liquid Filtration. West Conshohocken, Pennsylvania. https://www.techstreet.com/standards/ astm‐f838‐15ae1?product_id=2022158 (accessed 22 February 2019). Parenteral Drug Association. (Revised 2008). Sterilizing 5 Filtration of Liquids. PDA TR 26. Bethesda, Maryland. https://store.pda.org/tableofcontents/tr2608_toc.pdf (accessed 22 February 2019).

­  References

6 International Conference on Harmonization of

7

8

9

10

11

12

13

Technical Requirements for Registration of Pharmaceuticals for Human Use. (1999). Q5a(R1): Viral Safety Evaluation of Biotechnology Products Derived from Cell Lines of Human or Animal Origin. Tripartite Guideline. http://www.ich.org/fileadmin/Public_Web_ Site/ICH_Products/Guidelines/Quality/Q5A_R1/ Step4/Q5A_R1__Guideline.pdf (accessed 1 June 2018). Parenteral Drug Association. (2016). Consensus method for rating 0.1μm Mycoplasma reduction filters. PDA TR 75, Bethesda, Maryland. https://www.pda.org/ bookstore/product‐detail/3327‐tr‐75‐mycoplasma‐ reduction‐filters (accessed 22 February 2019). Center for Biologics Evaluation and Research, Food and Drug Administration. (1997). Points to consider in the manufacture and testing of monoclonal antibody products for human use. https://www.fda.gov/downloads/ BiologicsBloodVaccines/GuidanceComplianceRegulatory Information/OtherRecommendationsforManufacturers/ UCM153182.pdf (accessed 1 June 2018). Parenteral Drug Association (2008). Virus filtration. PDA TR 41 and PDA. J. Pharm. Sci. Technol. 62 (S‐4): 318–333. Lavoisier, A. and Schlaeppi, J.M. (2015). Early developability screen of therapeutic antibody candidates using Taylor dispersion analysis and UV area imaging detection. MAbs 7 (1): 77–83. Chen, D. (2014). Viral clearance using traditional, well understood unit operations (session I): virus‐retentive filtration. PDA J. Pharm. Sci. Technol. 68 (1): 38–50. Parenteral Drug Association. (2010). Preparation of virus spikes used for virus clearance studies. PDA TR 47, Bethesda, Maryland. https://store.pda.org/ TableOfContents/TR4710_TOC.pdf (accessed 22 February 2019). European Commission. (2008). EudraLex: The Rules Governing Medicinal Products in the European Union, Volume 4. EU Guidelines to Good Manufacturing Practice Medicinal Products for Human and Veterinary

14

15

16

17

18

19

Use, Annex I Manufacture of Sterile Medicinal Products. Brussels, Belgium. https://ec.europa.eu/ health/documents/eudralex/vol‐4_de (accessed 22 February 2019). The European Agency for the Evaluation of Medical Products, Human Medicines Evaluation Unit, CPMP. (1996). Note for Guidance on Manufacture of the Finished Dosage Form. London. https://www.ema. europa.eu/documents/scientific‐guideline/note‐ guidance‐manufacture‐finished‐dosage‐form‐first‐ version_en.pdf (accessed 22 February 2019). Parenteral Drug Association. (2014). Application of single‐use systems in pharmaceutical manufacturing. PDA TR 66. Bethesda, Maryland. https://store.pda.org/ TableOfContents/TR66_TOC.pdf (accessed 22 February 2019). Parenteral Drug Association (2005). Sterilizing filtration of gases. PDA TR 40. PDA J. Pharm. Sci. Technol 58 (1): 7–41. The United States Pharmacopeia. (2017). USP polymeric components and systems used in the manufacturing of pharmaceutical and biopharmaceutical products. Draft chapter. https:// www.gmp‐compliance.org/gmp‐news/usp‐draft‐ general‐chapters‐on‐plastic‐components‐and‐systems‐ used‐in‐the‐manufacturing‐of‐drug‐products (accessed 1 June 2018). Ding, W., Madsen, G., Mahajan, E. et al. (2014). Standardized extractables testing protocol for single‐ use systems in biomanufacturing. Pharm. Eng. 36 (6): 1–11. The European Medicines Agency. (2016). Draft Guideline on the Sterilisation of the Medicinal Product, Active Substance, Excipient and Primary Container. London. https://www.ema.europa.eu/ documents/scientific‐guideline/draft‐guideline‐ sterilisation‐medicinal‐product‐active‐substance‐ excipient‐primary‐container_en.pdf (accessed 22 February 2019).

141

143

11 Extractables/Leachables from Single‐Use Equipment Considerations from a (Bio)Pharmaceutical Manufacturer Alicja Sobańtka and Christian Weiner Octapharma Pharmazeutika Produktionsges.m.b.H, Vienna, Austria

11.1 ­Introduction Migration of leachables (chemical compounds that are released from a material into the contact medium) from primary packaging material into food has first been identi­ fied as a risk to the quality of food and customer safety in the 1980s. Successively, awareness of the risk from leacha­ bles to the end user expanded, it included pharmaceutical packaging and eventually, also processing materials such as filters, tubing, bags, and single‐use equipment. Leachables can pose a risk to human health. A direct health‐risk event can occur due to an overdose of a par­ ticular leachable following an acute and/or chronic toxi­ cological effect. Carcinogenic, mutagenic, teratogenic, and neurotoxic effects are particularly noxious. While for most toxicological effects a linear dose–response rela­ tionship applies, this is not true for substances with endo­ crine‐disruptive effects such as bisphenol A, for instance. Indirect health risk results from modification of the active pharmaceutical ingredient (API). Certain leach­ ables can interact with the API thus affecting the effi­ cacy and the activity of the latter. In some cases, adsorption to interfaces or interactions with leachables has caused protein aggregation or particle formation [1]. Molecules that can undergo a Michael addition reaction such as formaldehyde and acrylates, for exam­ ple, are known to modify therapeutic proteins [2, 3]. Leachables might cause an undesired increase in immu­ nogenicity [4]. Plastic material can adsorb the API [5, 6]. Leachables can also cause shift in pH due to the release of peroxides, for instance, an increase in viscosity due to discharge of dimethicones (polydimethylsiloxane), for example, and blocking of filters thus affecting the drug product (Figure 11.1). The (bio)pharmaceutical manufacturer should under­ stand any possible effects from potential leachables on the production processes and products. Potential interactions

with the API should be identified by appropriate ­p roduct specification tests and in long‐term stability studies. Today, more than 200 different plastic materials are listed. Over 7  500 monomers, 23  700 additives, and 13 000 pigments can be used to create tailor‐made poly­ mers [7]. New polymer formulations are being constantly developed. A polymer may contain a high, typically unknown number of ingredients and therefore, potential leachables. Extractables (chemical compounds that migrate from a material into a solvent under laboratory conditions) and leachables from plastic materials are generally additives such as antioxidants, processing aids like slipping agents, low‐number oligomers (n  =  1–5), residual solvents or polymer related monomers, secondary contamination from the packaging material of the wrapping plastic material, its washing and handling, and secondary and other ordinal number reaction products. Prominent examples are formaldehyde as a result of gamma irradia­ tion of plastic, the cell‐toxic bis(2,4‐di‐tert‐butylphenyl) phosphate, which is a breakdown product of the antioxi­ dant Irgafos 168®, the cell‐toxic 3,5‐dinitro‐bisphenol A from polycarbonate [8] and 7,9‐Di‐tert‐butyl‐1‐oxas­ piro[4.5]deca‐6,9‐diene‐2,8‐dione and 3‐[3,5]‐bis(tert‐ butyl)‐1‐hydroxy‐4‐poxocyclohexa‐2,5‐dienyl]propanoic acid from the antioxidant Irganox 1010 [5]. Such break­ down products may show different toxicological effects than their “mother‐molecule.” Migration of chemical compounds from plastic mate­ rial into any solid, pasty, liquid, or gaseous contact medium is a complex process that is at the same time kinetically and thermodynamically driven. The kinetic and thermodynamic drivers are diffusion and solubility, respectively. Occurrence and extent of leachables thus depend on numerous parameters such as contact time, temperature, solubility of the leachable in both the

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

11  Extractables/Leachables from Single‐Use Equipment

There are exceptions to the migration process that have been described before. Some extractables/leacha­ bles appear only in the beginning of contact with the accordant medium, in the beginning and in the end, only in the halftime, or only at the end of the contact time [9]. The reason therefore is that extractables/leach­ ables can undergo chemical reactions with the contact medium, atmospheric oxygen, light, and/or among each other thus transforming to other molecules with differ­ ent release behavior.

Leachables

Toxicity

API modification

Direct

Health risk

11.2 ­Regulatory Environment

Indirect

Figure 11.1  Direct and indirect health risk from leachables. Migration of leachables over time Concentration leachable

144

Temperature 1 Temperature 2 > Temperature 1

6–12 h

Contact time

Figure 11.2  General illustration of migration of leachables in the course of time.

­ olymer and the contact medium, mobility (viscoelastic­ p ity) of the polymer, size and structure of the leachable, and the free volume of the polymer (space that is not occupied by the molecules). Migration of extractables starts after an initial contact time that can vary between a few minutes to a couple of hours. The concentration of the extractable/leachable increases over time, however, not linearly. After some time, usually 6–12 hours, the equilibrium state is reached where there is hardly any or no more increase in the con­ centrations of the individual extractables. Note that the time to reach equilibrium can differ for plastic materials with different preceding aging conditions. In general, the  longer the use of the material and the higher the ­temperature, the higher the concentration of released extractables/leachables and therefore, the higher the probability of inserting unwanted leachables into the drug product (Figure 11.2). Moreover, the risk increases if the composition of the product has a high extraction force, for example alcohol or fat. Surfactants such as poly­sorbate may have an impact on extraction of chemi­ cal compounds from plastic materials.

To date, there is no standardized mandatory regulation of single‐use equipment. In general, however, the Code of Federal Regulations (CFR) of the United States obliges the (bio)pharmaceutical manufacturer to assure that “all equipment surfaces that come in contact with products be free of leachable contaminants that will hasten the deterioration of the product or otherwise render it less suitable for the intended use” [10]. In addition, plastic materials destined for use in food‐ and pharmaceutical industry are regulated to some extent, which is discussed in the following sections. 11.2.1  Pharmacopeia Chapters Polyethylene (PE), polyethylene terephthalate (PET), polypropylene (PP), and elastomers are regulated as container closures [11, 12]. U.S. Pharmacopeia’s chap­ ter  661 (USP ) requires physicochemical tests that suggest chemical safety assessment by restricting the amount of extractable heavy metals and the extract­ able nonvolatile residue [11]. Accordant regulations are offered by the European Pharmacopeia (EP) [13–18]. USP is of limited value because it is applicable to only three different pure polymers and to a sum parameter (nonvolatile residue) for organic extractables. In reality, plastic materials that are used in pharmaceuti­ cal manufacturing are often composed of various poly­ mers and/or copolymers. Extractables/leachables can be volatile, semi‐volatile, and nonvolatile in nature and their tolerable amount highly depends on their chemical ­ nature. While 1 mg of erucamide might not pose any risk to patients upon exposure, 1 mg of bisphenol A, for example, can cause severe damage to human health (Section 11.3.2). In addition to the requirements with respect to extract­ able heavy metals and nonvolatile residue, USP asks for determination of acidity/alkalinity, reducing substances, volatile sulfides, and extractable ammonium from elastomers [12]. While the aforementioned param­ eters are useful for evaluation of the overall suitability of

11.2  Regulatory Environment

the elastomeric closure, they are not sufficient for a thor­ ough chemical safety assessment. Accordant regulation is offered by the EP 3.2.9 [19]. EP 3.1.9 offers tests spe­ cific to silicone, which is widely used in pharmaceutical manufacturing in the form of tubing [20]. Elemental impurities are regulated in accordance with the International Council for Harmonisation (ICH) of Technical Requirements for Pharmaceuticals for Human Use [21] and the U.S. Pharmacopeia [22, 23]. USP and offer guidance on assessment of extractables and drug product leachables associated with pharmaceutical packaging and delivery systems including preparation of extracts, extraction conditions, and characterization of the extracts  [24, 25]. While the (bio)pharmaceutical manufacturer is advised to respect both USP chapters for extractables characterization from primary packaging, there is nothing wrong with making use of the chapters for extractables characterization from other plastic processing material. Extractables characteri­ zation according to USP is widely accepted. 11.2.2  Biological Reactivity and Chemical Safety Typically, all components of single‐use equipment that are dedicated for pharmaceutical manufacturing fulfill requirements of USP class VI. An USP class VI plastic material shows no obvious local or systemic responses from a living system or tissue. For USP class testing, dif­ ferent extracts (sodium chloride, alcohol saline, polyeth­ ylene glycol, and vegetable oil) of the plastic material are generated and subsequently intravenously and intraperi­ toneally administered into a test model. For USP class VI, the plastic material is additionally implanted. For sys­ temic injection, the test model is observed for three days for signs of toxicity or death. For intracutaneous reactiv­ ity, the test models are scored at 24, 48, and 72 hours to determine whether a significant reaction occurred. For muscle implantation, the test model is evaluated after five or seven days to determine whether a significant reaction occurred. If no reactions are observed, the plas­ tic material is considered safe. Thus, overall biocompat­ ibility can be regarded as an indicator for chemical safety. Biocompatibility testing, however, does not necessarily imply chemical safety. Plastic material can fulfill USP class VI requirements and yet not be chemically safe for a particular application due to the release of a toxicologi­ cally critical amount of an extractable/leachable. Examples of biocompatible but not chemically safe plas­ tic materials are: (i) USP class VI polyvinylidene fluoride (PVDF) connectors that leached around 15 μg of the critical 1H,1H,2H,2H‐perfluorooctane sulfonic acid (Chemical Abstracts Service [CAS] No. 27619‐97‐2) per gram of connector; (ii) USP class VI silicone tubing that discharged 30 μg of ­mutagenic tetrachlorinated aromatic

compounds per cm2 of tubing; (iii) a USP class VI filter (10″) that released 40 mg of diphenyl sulfone; and (iv) medical‐grade USP class VI polyvinyl chloride that leached phthalates (CAS No. 85‐68‐7 : 1,2‐benzenedicar­ boxylic acid, butyl phenylmethylester phthalate [BBP], CAS No. 84‐74‐2 : 1,2‐­benzenedicarboxylic acid, dibutyl ester phthalate [DBP], CAS No.117‐81‐7 : 1,2‐benzenedi­ carboxylic acid, bis[2‐ethylhexyl]ester phthalate [DEHP, DOP]) of very high concern according to the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) in addition to various chlorinated organic compounds (all data confidential). 11.2.3  “Pharma Grade,” “Medical Grade” Plastic for (bio)pharmaceutical manufacturing is some­ times supplied in the so‐called “pharma grade.” At this stage, it should be mentioned that an official definition of “pharma grade” does not exist. “Pharma grade” usually refers to tracking of changes in the supply chain of mate­ rials that are used for production of the target plastic. A legal definition of “medical grade” to date does not exist either. Medical grade may refer to individual aspects of medical device compliance testing such as the International Organization for Standardization (ISO) 10993 series and most prominently, to biocompatibility testing. Single-use equipment however, is not a medical device and therefore, its qualification is not performed under the legal framework for medical devices, which is different and independent from the qualification of pharmaceutical processing materials. 11.2.4  Code of Federal Regulations – Food Grade The CFR manages additives to specific polymers that are intended for the use in food manufacturing [26–36]. For oral administration of drug products, CFR certification can be quite useful as support to chemical safety assess­ ment. Accordant food safety regulation can be found in the European Union (EU) [37, 38]. For drug products that are applied via other routes of administration such as par­ enteral or inhalation, food‐grade plastic is of limited value. The tolerable intake (TI) of a chemical substance is typi­ cally much higher when administered orally in compari­ son to intravenously, for instance. In other cases such as mercury, for example, oral uptake is of little relevance since the element is excreted without much interaction in the human body. Intravenous absorption can, however, cause damage to the kidneys. Due to the absence of a legally binding definition of “pharma‐ and/or medical‐ grade” plastic that would ensure chemical safety, the (bio) pharmaceutical manufacturer often makes use of food‐ grade plastic. In fact, food‐grade certificates are often requested by authorities.

145

11  Extractables/Leachables from Single‐Use Equipment

11.2.5 REACH The (bio)pharmaceutical manufacturer should verify REACH compliance of the single‐use equipment in order to exclude introduction of high‐risk leachables into the drug product. European companies must identify and manage the risks linked to the substances they manufac­ ture and market in the EU. The responsibility for fulfill­ ing the requirements of REACH, such as pre‐registration or registration lies with the importers established in the EU, or with the representative of a non‐EU manufacturer established in the EU. 11.2.6  Regulatory Responsibility Chart Although it is the (bio)pharmaceutical manufacturer who is ultimately responsible for suitability of the single‐ use equipment for its intended use, the single‐use equip­ ment supplier and his sub‐suppliers are not free of responsibility for regulatory compliance of the individual single‐use equipment materials and parts (Figure 11.3). Plastic resin and plastic component suppliers as well as the single‐use equipment supplier should provide raw materials, plastic parts, and ultimately, single‐use equip­ ment that are in accordance with the requirements as described in the different Pharmacopeia chapters, biologi­ cally not reactive, and wherever possible, free of hazardous chemical substances. Ideally, plastic component suppliers also provide generic extractables data. Extractables testing of polymer resins can be indicative of the extractables profile of the compound [39].

Regulatory level

146

11.3 ­The (Bio)Pharmaceutical Manufacturer’s Approach 11.3.1  Risk Mitigation 11.3.1.1  Chemical Compatibility

Prior to extractables profiling, chemical compatibility/ resistance of all single‐use equipment components with the respective product stream(s) must be verified under worst‐case production conditions regarding operating time, temperature, and pressure. Limited or poor chemi­ cal compatibility will attack the plastic material, thus potentially releasing a much higher amount of leach­ ables. Note that for sterile‐used single‐use equipment, chemical compatibility must be checked after steriliza­ tion. The reason therefore is that sterilization methods such as autoclave, gamma irradiation, X‐ray, ethylene oxide (EtO), and e‐beam stress the plastic material, thus possibly altering its physicochemical parameters. For example, the energy‐rich gamma irradiation triggers chemical reactions within the polymer, changing its chemical profile compared to the untreated material. This might also affect chemical compatibility. Effects of sterilization on plastics and elastomers are extensively discussed in [40]. Information about chemical resistance and compatibility with sterilization methods can usually be obtained from the single‐use equipment supplier. In the latter case, the (bio)pharmaceutical manufacturer should verify if in the context of compatibility testing, the chemical profile has been checked.

• Pharmacopeia • USP

class VI

• REACH

• Chemical

safety • Pharmacopeia • USP

class VI

• CFR

• REACH

• Extractables

(optional) • REACH

• Pharmacopeia

(optional) (optional) • USP class VI (optional) • Extractables (optional) • CFR

• REACH

• Pharmacopeia

(optional) (optional) • USP class VI (optional) • Extractables (optional) • CFR

(Bio)Pharmaceutical manufacturer

Single-use equipment supplier/manufacturer

Plastic component supplier/manufacturer

Figure 11.3  Responsibility chart regulatory compliance of single‐use equipment.

Plastic resin supplier/manufacturer

11.3  The (Bio)Pharmaceutical Manufacturer’s Approach

Figure 11.4  Pharmaceutical manufacturing process including low‐ and high‐risk steps with regard to leachables from plastic materials.

diafiltration steps enhance this clearance [41]. Ultra‐ and diafiltration are recognized for reducing the risk from leachables [42, 43]. The purification capacity of processing steps can be regarded as indicative of the leachables removal power. Caution must, however, be exercised as to the removal capacity of potential clearance steps. The clearance power might not amount to 100%. The (bio)pharmaceutical manufacturer is therefore advised to carefully review and prove the leachables removal capacity of potential clear­ ance steps for example, by means of appropriate studies. If a single‐use equipment is used at a stage of the phar­ maceutical manufacturing process where no further leachables clearance can be expected, it must be assumed that any chemical compound that migrates from the individual parts of the single‐use equipment ends up in the final drug product. An example for such high‐risk single‐use equipment is a filling set. 11.3.1.3 Pre‐Flush

Often and especially, when filtration is involved, pre‐ flushes either with water for injection, buffer, or solvent are

Leachables from plastic processing materials

Upstream last clearance step, low risk

Single‐use equipment can be used at different stages of a pharmaceutical manufacturing process. The risk from leachables from single‐use equipment for the efficacy and safety of the final product can vary depending on at what stage of processing it is used (Figure 11.4). Certain processing steps with purification capability show a substantial leachables removal capacity and can hence be regarded as clearance steps (compare also Chapter 17). Chromatographic purification in the form of affinity, ion exchange, and size exclusion chromatog­ raphy is widely used in pharmaceutical manufacturing. Among the different methods, affinity chromatography, where the API is bound to the chromatography resin while impurities are washed off, is the most powerful one to remove leachables from the API. Note that this is only true for chemical compounds that are not physically or chemically bound to the API. If the API is a large molecule in comparison to typical leachables (MW up to 1000 g/mol), it is retained by ultra­ filtration while potential leachables pass to the permeate thus being removed from the drug product. Subsequent

Downstream last clearance step, high risk

11.3.1.2  Clearance Steps

API production

Purification

Formulation

Fill finish

Clinical application

Removal of leachables from the drug product

147

148

11  Extractables/Leachables from Single‐Use Equipment Chemical compound in the plastic material bulk

Chemical compound in the plastic material surface

It takes time for the compound to migrate to the surface

Plastic material

Figure 11.5  Illustration of the local distribution of potential leachables and other impurities in and on plastic material. Gray spheres can be removed by pre‐flush.

carried out before processing the drug product. While such pre‐flush can be helpful for the performance of the processing step (filter wetting for optimal filtration perfor­ mance), it can also flush any residual impurities and leach­ ables from the product‐contacted surface of the plastic material (for example, bloomed additives [44]). The vast majority of extractables can, however, not be removed by pre‐flush. The reason therefore is that most extractables are located in the bulk of the plastic material rather than at its surface (Figure 11.5). They are continuously released in the course of contact between the plastic material and the respective medium. Pre‐flush will therefore be more ben­ eficial for short‐ than for long‐term processing (Figure 11.2). Analysis of a sterile filter that would typically be used in a single‐use equipment revealed that PP and oxidized Irgafos 168 are discharged from the membrane (hydro­ philic PVDF), however, depending on the flush medium: water and other aqueous solutions would not remove any chemical compounds (confidential data). These find­ ings did not match the supplier’s recommendation to pre‐flush the filter with water for removal of organic matter expressed by means of the unspecific sum param­ eter total organic carbon. In another case of a sterilizing filter containing a PES membrane, it was found that large amounts of caprolactam were washed off the filter by aqueous media (confidential data). The caprolactam was confirmed as an integral ingredient of the filter manufac­ turing process. Such readily flushable substances could result in a heightened level of extractables/leachables in the containers filled at the beginning of a continous fill­ ing process. 11.3.2  Chemical Safety Assessment The challenge to predefine those leachables that would migrate from a plastic material into the drug product and the complexity of the migration process prevent a paper/ computer‐based estimation of leachables in the final

drug product with reliable accuracy. Therefore, testing of the plastic material and, in worst case, the final product is required (Figure 11.6). 11.3.2.1  Extractables Profiling

Every plastic component of a single‐use equipment will  contribute with individual types and amounts of extractables/leachables. In a first step, a controlled extractables study should be performed to get an idea of the types of chemical compounds that might leach from the plastic material. Some suppliers and sub‐suppliers offer generic extractables data. The appropriateness of such extractables data for chemical safety assessment of a single‐use equipment must be carefully reviewed by the (bio)pharmaceutical manufacturer. A controlled extractables study looks for the yet unknown chemical compounds that might be released from a plastic mate­ rial. It thus gives primarily qualitative information about the types of chemical compounds (“what might migrate from the plastic?”) and secondly, due to the use of refer­ ence substances, semiquantitative data (“how much might migrate?”). A key point in establishing conditions for a controlled extraction is the question about the proper amount of extractables that should be released. There are three possibilities: firstly, extraction of the total amount of an extractable present in the polymer material (exhaustive extraction). Exhaustive extraction is typically used for determination of residues such as EtO (sterilization residue) on view of the residue’s specified limit and for critical compounds such as diisocyanate in polyurethane or nitrosamines. Secondly, extraction of the absolute amount of an extractable that can be extracted from the material. This case ideally depicts a worst‐case spectrum and accumulation of potential leachables in the drug product (controlled extraction). Thirdly, extraction of the absolute amount of an extractable that will poten­ tially accumulate in a drug product during processing and/or storage (simulated‐use extraction). U.S. Pharmacopeia [24, 25], the Product Quality Research Institute (PQRI) [45], and the BioPhorum Operations Group (BPOG) [46] offer guidance on how to perform extractables studies. Extractables characteri­ zation according to PQRI and the standardized BPOG extractables testing protocol are widely accepted. The BPOG protocol is very comprehensive using samples from different lots and detailing the extraction proce­ dure. Extracts after different time spans are prepared and characterized by appropriate analytical methods. The downside of the BPOG protocol is the expense in terms of scope and costs, and use of questionable solvents such as 1% polysorbate 80, which is known to cause interfer­ ences with the anticipated analytical methods and 5 M NaCl, which is of very little if no additional value in com­ parison to extraction in water [46].

11.3  The (Bio)Pharmaceutical Manufacturer’s Approach

Chemical safety assessment

Extractables profiling

Exposure calculation

Toxicological risk assessment One or more extractables significantly exceed tolerable intake

Discard material

One or more extractables moderately exceed tolerable intake

One of more extractables significantly exceed tolerable intake

Simulated-use extractables study

No extractable exceeds tolerable intake

No extractable exceeds tolerable intake

Single-use equipment chemically safe

One or few extractables slightly exceed tolerable intake

One or more leachables exceed tolerable intake

Leachables study

No leachable exceeds tolerable intake

Figure 11.6  Proposed workflow for chemical safety assessment.

Parameters of an extractables study should in any case respect the use of the single‐use equipment. Therefore, rather than following standardized protocols, it is impor­ tant to make an optimal choice of relevant parameters associated with extractables studies. 11.3.2.2  Controlled Extractables Study

A controlled extractables study overestimates the types and amounts of potential leachables (worst‐case leacha­ bles profile). It is the safest choice to start within an extractables and leachables assessment. Important aspects of a controlled extractables study are shown in Table 11.1. An extractables study can be either performed for the entire single‐use equipment and/or its individual components. The advantage of the latter “building block” approach is that extractables can be assigned to the individual parts. In the case of an extractable of toxicological concern, the component can hence easily be replaced. This approach is initially more cost‐inten­ sive. However, if the single‐use equipment supplier or the sub‐suppliers change the formula, the production

process, or the production site of a single‐use equip­ ment component thus requiring anew extractables characterization, only the material costs for the indi­ vidual plastic component are spent. Another advantage of extracting individual single‐use equipment compo­ nents is better detection of extractables due to less peak overlap, which can occur when a high number of extractables is present. 11.3.2.3  Sum Parameters

In the past, extractables were often characterized by means of sum parameters such as the nonvolatile residue (NVR) or total organic carbon. Therefore, the material was extracted in different solvents. Subsequently, the NVR was generated by evaporating the solvent and gravimetrical determination. Often, an infrared spec­ troscopy (IR) spectrum of the NVR was measured and interpreted with the help of knowledge about the poly­ mer formulation. The risks of using sum parameters such as the NVR are missing volatile and semi‐volatile extractables and impeding a toxicological assessment.

149

Table 11.1  Checklist for extractables studies. Laboratory

Extractables (and leachables) studies should be performed at qualified and experienced laboratories. The (bio) pharmaceutical manufacturer is dependent on the expertise of the chosen laboratorya and should therefore not hesitate to take the time to thoroughly interview the laboratory staff on details of the extractables study. An initial technical visit followed by a quality audit of the laboratory and all its sites where measurements are performed is highly recommendable.

Samples

Material samples from different lots should be used (where possible) to cover potential lot‐to‐lot variability.

Reporting limits/AET

In extractables studies for primary packaging, the AET is the concentration above which extractables should be identified and reported in the toxicological assessment. The AET is calculated under consideration of the applicable safety concern threshold (see the TTC concept 11.3.2.8) and the product dose in addition to the analytical uncertainty. Therefore, clinical application of the accordant drug product(s) must be known. If the calculated AET is lower than the technically possible detection and quantification limits, any detected extractables can be reported. It is generally advisable to report any extractables above the limit of detection (LOD)/ limit of quantification (LOQ). This way, the extractables study does not need to be repeated if the application of the material changes thus leading to a lower AET.

Model solvents

Polar (water, alcohol such as ethanol or propanol or aqueous alcoholic solutions) and nonpolar (hexane) solvents as well as acidic and alkaline buffers can be used to cover hydrophilic and hydrophobic respectively, polar and nonpolar extractables. The appropriate choice of model solvents should always reflect the chemical properties of the relevant drug product(s).

Temperature

Room temperature to 40/50 °C is commonly used for extraction depending on the processing/storage temperature. Note that the higher the temperature, the faster the migration and therefore ultimately, the higher the amount of extractables.

Time

Typically, extraction of plastic material is carried out for 24 hours. If the operating time, however, extends to seven days, for instance, the extractables study should be performed for at least seven days. In certain cases it might be relevant to check the release of extractables after different periods for example, in the case of filling in continuous mode (verification of the distribution of extractables among the filled containers).

Accelerated extraction

If a single‐use equipment is used for weeks or months, extraction can be accelerated to accomplish the goal of generating an appropriate extractables profile within reasonable timeframes. Acceleration can be achieved by means of statistical and physical effects, solvent selection, and temperature. Statistical effects include, for example, exaggeration of the conditions of the test article contact such as increase of contact surface to solvent ratio and increase of the number of extraction cycles. Physical effects refer to the use of external physical force such as agitation and sonication. Note that sonication energy cannot be easily controlled. Solvents can be chosen that swell the polymer, thus allowing penetration of the polymeric structure and therefore, facilitating migration of extractables. It is important that the extractables be well soluble in the selected solvent. If this is not the case, insufficiently high concentrations of the extractables would be obtained, failing to present a complete extractables profile. Increase of temperature has a dual effect by both elevating the kinetic energy of the extractable and decreasing the viscosity of the matrix. Higher temperatures reduce interactions among the matrix molecules, thus increasing the extractables flux. The temperature should not be too high though. Extraction temperatures close to the polymer’s melting temperature may change the polymeric intrinsic structure. For many extractions such as Soxhlet and reflux, extraction below the boiling point of the solvent mixture is the best choice. Soxhlet extraction is the best technique to favorably control the concentration gradient between solvent and material and it can easily be validated. Reflux involves considerable temperature acceleration.

Analytical methods

Orthogonal methods should be used to capture any possible extractables. Chromatographic methods such as gas- and liquid chromatography are used to separate the different volatile, semi‐volatile, and nonvolatile chemical compounds. Different detectors (UV, flame ion detector, diode array detector detect the various polar and nonpolar compounds. Additional identification and semi‐quantification is performed by means of mass spectroscopy that is typically coupled with the chromatographic methods. Derivatization can be helpful for detection of species that may otherwise not be sensitive enough in regular measurement. An example is the derivatization of fatty acids with N,O‐bistrifluoroacetamide to obtain the accordant trimethylsilyl ester. Inductively coupled plasma is used for screening of elemental impurities. Gas chromatography‐thermic energy analysis can be applied to screen for nitrosamines from rubber elements. Ionic chromatography conductivity detector can be used for detection of anions. Atomic absorption spectroscopy is suitable for identification of silicone Gel permeation chromatography would detect larger molecules (molecular weight beyond 3000 g/mol for example, polymer degradation products). Quadrupole time‐of‐flight and tandem mass spectroscopy experiments enable determination of the exact mass of a molecule and a meaningful mass fragmentation pattern, which can be exploited for identification of chemical compounds. Typical ionization modes are atmospheric pressure chemical ionization and electrospray ionization (ESI). High‐performance liquid chromatography (HPLC)‐ESI is appropriate for detection of PET extractables and perfluorinated substances. The latter can be released from fluorinated polymers [47]. Analytical methods do not need to be fully validated for extractables studies. However, minimum quality criteria such as method sensitivity, selectivity (capability to separate chemical compounds), LOD, LOQ, analytical uncertainty, linear calibration curve based on different and appropriate reference standards for semi‐quantification, and recovery experiments should be assessed to demonstrate method suitability and to provide information about the reliability of the data. Blanks must be run in order not to mistake chemical compounds from laboratory equipment and materials for extractables. Numerous set points are to be defined for each analytical method. Choice of modules, chromatography columns, and temperature programs must be compromised to cover a broad variety of yet unknown chemical compounds that can be expected in an extractables study. If necessary, specific set points can be chosen in a follow‐up study.

11.3  The (Bio)Pharmaceutical Manufacturer’s Approach

Table 11.1  (Continued) Reference standards

A sufficiently wide range of reference standards should be selected to cover the chemical nature of possible extractables. If certain extractables can be expected, they should be added as authentic reference standards. If an authentic reference standard is not available, an appropriate surrogate can be used.

Surface area/ An extractables study should be designed in such a way that potential extractables occur in sufficiently high mass to solvent concentrations in order to be detected. Ratios ranging from 1 cm2/ml [42, 43] to 6 cm2/ml or 0.2 g/ml [48] are volume ratio recommended. Sample If concentration of the extracts is required (for example, to achieve a defined surface area/mass to solvent volume concentration ratio), the concentration factor should not exceed 10 in order to avoid falsification (loss or gain of chemical compounds that are not material‐related extractables). Either way, the original extracts should be also analyzed. Pretreatment

Sterilization (autoclave, gamma irradiation, EtO, e‐beam, and X‐ray) and pre‐flush in production and the sequence of the different pretreatment steps can stress the material or conversely remove chemical compounds from the surface of the plastic material and should therefore be considered prior to extraction of the material.

a

 Regarding knowledge of the polymer, expectable extractables, analytical methods, choice of reference standards, detectors, and ionization mode(s), for instance.

Analysis of extractables from different materials, deter­ mined by means of the sum parameter NVR and IR on the one hand and, by orthogonal analytical methods on the other hand (Table 11.1) revealed that there is hardly any or only a poor correlation regarding the amount and the types of extractables (confidential data). In the example of a filter, 54.7 mg of NVR was detected after extraction in water. The NVR was assigned to poly(acrylate)ester after IR measurement. A controlled extractables study on the same filter showed a total of only 11.4 mg of extractables and no poly(acrylate)ester. Instead, fatty acids, antioxi­ dants, and siloxanes were identified. In the case of a sili­ cone tubing, the NVR was significantly lower than the total amount of extractables found by orthogonal analytical methods. The IR spectrum of the NVR classified silicone oligomers. Gas chromatography/mass spectroscopy (GC/ MS) and liquid chromatography/mass spectroscopy (LC/ MS), on the other hand, found five different siloxanes, tri­ methylsilanol, and polypropylene glycol‐related com­ pounds in addition to silicone oligomers [49]. IR measurements are not suitable for analysis of a mix­ ture of compounds because the likelihood of peak overlap is quite high. Many extractables carry the same functional groups thus giving similar absorbent patterns, thereby preventing identification of individual extractables. 11.3.2.4  Unknown Compounds

In almost every extractables study, chemical compounds whose identity or structure cannot be assessed with suf­ ficient reliability occur in toxicologically relevant con­ centrations. The mass spectrometric fragmentation pattern might not match any library or literature spec­ trum, it might not be possible to assign it unambigu­ ously to a chemical substance, the molecular weight might hint too many different possible compounds, the chromatographic retention index and the mass spec­ trum might not match any authentic chemical, or the extractable might not even yield a mass spectrum.

In this case, patient exposure (for calculation of the patient exposure see Section 11.3.2.7) to the unknown extractable should not exceed the applicable threshold of toxicological concern (TTC) [50]. If not toxicologically rel­ evant, no further investigation of the unknown compound needs to be performed. However, if the unknown com­ pound appears in a concentration of toxicological concern, additional measures must be undertaken. Supplementary analytical investigation might be appropriate for further identification. Nuclear magnetic resonance is a very pow­ erful analytical method, but it requires sufficiently high concentration of the analyte (>100 μg/ml, an unusual high concentration for extractables/leachables). It can be much more expedient to contact the supplier and/or the manu­ facturer of the respective single‐use equipment compo­ nent and ask for support. Usually, the unknown extractable can be assigned to an additive or processing aid or one of its known degradation products. The manufacturer or the supplier may also have more in‐depth information about potential extractables from his plastic material. Alternatively, if sufficient information about the unknown compound is present, Cramer classification can be performed to classify and rank extractables according to their expected level of oral systemic toxicity. Therefore, extractables are categorized into three differ­ ent classes indicating low (i), middle (ii), and high (iii) level of concern. Each class is associated with a human exposure level, below which the compounds are consid­ ered to present a negligible risk to human health [51]. Although the Cramer‐derived human exposure level is strictly applicable for oral administration, it can be indica­tive for other routes of administration. Guidance on how to deal with unknowns is provided in Figure 11.7. 11.3.2.5  Simulated‐Use Extractables Study

If an extractables study yields unfavorable results for a plastic component, a simulated‐use extractables study can be carried out under more process‐like conditions to

151

152

11  Extractables/Leachables from Single‐Use Equipment

Unknown extractable

Concentration of unknown below TTC

Concentration of unknown above TTC

Unknown acceptable Determination of accurate mass

Sufficient information

Not sufficient information

Support from supplier

Sufficient information

Unknown identified

Unknown cannot be identified

Possible

Additional analysis

Identification possible

Identification not possible

Cramer classification

Not possible Take the risk or discard material

Figure 11.7  Guidance on how to deal with unknown extractables.

confirm or reject the results from the initial extractables study. Therefore, analytical methods, set points, and ref­ erence standards can be adapted accordingly. If, how­ ever, the number and/or the amount of extractables of concern is very high and it can be assumed that it would not be significantly lower in a simulated‐use extractables study, the (bio)pharmaceutical manufacturer might con­ sider discarding the plastic component and replacing it by another one with an acceptable extractables profile. Skipping an initial comprehensive extractables study in favor of a simulated‐use extractables study is not recom­ mendable because stand‐alone, such studies bare the risk of missing extractable compounds, for example, because they are not present in the extracts in sufficiently high concentration in order to be detected.

11.3.2.6  Leachables Study

If the results of either the initial extractables study or the simulated‐use extractables study are still unacceptable, the real concentration of the extractables of concern can be verified by means of leachables studies in the accord­ ant drug product(s). Leachables studies (see also Chapter  17) require fully validated analytical methods, thus requiring significant work‐ and cost load. In some cases, leachables studies are not useful or even possible, for example, if the drug product matrix is very complex thus impeding reliable measurement. Hence, suitability testing should be performed prior to a leachables study. In other cases, an identified extractable would never be detected in a drug product, for example, in the case of acrylic acid in a protein‐based drug product because

11.4  The (Bio)Pharmaceutical Manufacturer’s Challenges

the acrylic acid would immediately react with the pro­ tein to form another species. It might then be more fea­ sible to specify the tolerable extracted limit of the target compound from the polymer material. In some leacha­ bles studies it might be necessary to first remove the API, however, without risk of losing the potential leach­ able. The costs for the leachables studies can exceed a reasonable level. It is not unusual to end up with hun­ dreds of different extractables and therefore, potential leachables in one drug product. In this case, it might be the preferable option to discard the material. 11.3.2.7  Exposure Scenario

Once extractables profiling of the single‐use equipment or its individual components has been carried out, the extractables amount in the final drug product needs to be estimated. Therefore, extractables raw data have to be converted to an appropriate unit such as μg per single‐ use set, total product contacted surface area, or per part. The amount of every extractable is then divided by the minimum regularly processed bulk volume in accord­ ance with a worst‐case approach (highest concentration of extractables). The extractables concentrations are subsequently mul­ tiplied by the human dose of the accordant drug product(s) according to clinical application (for some drug products, single- and chronic administration have to be distinguished). 11.3.2.8  Toxicological Risk Assessment

The calculated theoretical patient exposure can now be toxicologically assessed. Note that depending on the drug product, patient exposure to extractables can be different. Therefore, extractables from a material can be acceptable for a drug product but not for another one, which is administered in a higher amount, for instance. Toxicological risk assessment is performed twice. In the first row, the extractables from the single‐use equip­ ment are evaluated. Secondly, the extractables from all relevant plastic materials including the single‐use equipment are assessed (Figure  11.4 downstream pro­ cessing after the last clearance step and clinical devices). The reason therefore is that certain plastic materials release the same types of chemical compounds, which would ultimately accumulate in the drug product thus yielding a higher concentration than the one from the single‐use equipment only. While the concentration of, for example, stearic acid from the entire single‐use equipment might be tolerable for the clinical use of an accordant drug product, the sum of stearic acid from all plastic materials might reach levels that suggest a risk to the patient. Firstly, patient exposure is compared to the applicable TTC [50]. Thereby, the TTC is an acceptable limit even for mutagenic impurities in drug products, and represents

a worst‐case scenario (daily exposure to a mutagenic compound). According to the ICH guideline M7,  no further risk assessment has to be conducted for com­ pounds that do not exceed this threshold. For all sub­ stances that exceed the TTC, toxicological information has to be obtained. Therefore, a TI is derived based on the most appropriate available toxicological informa­ tion, respecting the route of administration and the test species by introduction of accordant uncertainty fac­ tors. The uncertainty factors are used to convert the raw toxicological data into tolerable intakes that con­ sider the individual application taking into account the intra-individual variation between humans, the  differ­ ent species (toxicological data that was generated based on a mouse can be extrapolated to humans), the various routes of administration, and the toxicological end­ points. The TI is compared to the patients’ exposure to the respective extractables. A chemical compound is considered as safe for the patients’ health if its exposure to patients is lower than the TI. Some extractable compounds are listed in official guidelines, e.g. the ICH guideline Q3C for solvents [52], the ICH guideline Q3D for elemental impurities [21], or the EMA guideline for metals [53]. If acceptable intakes (permitted daily exposure) are listed in a guideline, the exposure of patients is compared to the TIs derived from these values. Several very similar or even identical com­ pounds, for example, linear dimethicones and certain fatty acids (C12–C16 saturated fatty acids, for example), are summed up under a general designation assuming the same impact on the human body. If no toxicological data are available for a specific chemical compound, appropriate surrogates can be used assuming the same or a very similar mode of action in the human body. If the ensemble of all relevant extractables does not bare any toxicological risk for the patient treated with the accordant drug product(s), the single‐use equipment can be regarded as chemically safe and suitable for its intended use.

11.4 ­The (Bio)Pharmaceutical Manufacturer’s Challenges 11.4.1  Supply Chain of Single‐Use Equipment Single‐use equipment is typically composed of different plastic materials such as tubing, connectors, sensors, bags, filters, and filling needles (classical stainless steel needles may contain some additional polycarbonate, for instance). Each component will most probably come from a different supplier bringing along its own individ­ ual supply chain history. Every manufacturing step requires different chemicals and bares the  potential of carryover of unwanted impurities into the subsequent

153

11  Extractables/Leachables from Single‐Use Equipment

Chemicals, storage stabilizers Monomer synthesis

Catalysts, stabilizers, antioxidants, processing aids,... Polymer manufacturing

Stabilizers, antioxidants, processing aids,... Masterbatch

Lubricants, colorants,...

Plastic component

154

Molding, dimensioning

Figure 11.8  Illustration of the supply chain of a plastic component that forms part of a single‐use equipment.

manufacturing step and eventually, into the final plastic component (see Figure 11.8). Figure 11.8 illustrates the complexity of the multilevel supply chain of a plastic component, which eventually becomes part of a single‐use equipment. Connectors, tubing, and other small parts of a single‐use equipment are rather simple materials usually made of one single polymer material. Bags can be composed of multiple lay­ ers made of different polymer materials. Often the bags’ ports are made of yet another polymer material from a different supplier. Filters are by far the most complex materials consisting of a membrane, filter support, cap­ sule, end caps, and O‐rings. Hence, one filter can have up to five different polymer suppliers. Production of each polymer requires accordant resins, additives, processing aids, solvents, and other chemicals. Polymer resins just like additives, processing aids, and other chemicals that take part in the manufacture of plastic are usually produced under a less‐stringent quality system than the final single‐use equipment that is anticipated for pharmaceutical production. In consequence, the quality of a polymer can vary from lot to lot. Differences in the extractables profiles of different lots of a certain polymer material are usually observed [54]. This handicap can be overcome by using samples from different lots for extractables characterization. The same polymer can, depending on the supplier, be  different in nature due to its distinct formulation. Therefore, the PP of one supplier will probably be differ­ ent from the PP of another supplier [54]. As a result, if a plastic component is exchanged by another one made of the same polymer, chemical safety assessment of the ­single‐use equipment has to be reviewed. The overall complexity makes it obviously very diffi­ cult to control the supply chain of a single‐use equip­ ment. While business relations among the different suppliers are regulated by means of supplier‐ and qual­ ity agreements, there are restrictions in the information flow. No producer will reveal his “recipe” and therefore, knowledge about the plastic material declines from one

level of the supply chain to the next. The single‐use equipment supplier might hence not even be aware of the exact materials of construction or not communicate them accordingly. Examples are a PVDF filter mem­ brane that is actually a PVDF–polyacrylate copolymer and a PP connector that is really made of PE–PP copoly­ mer. The (bio)pharmaceutical manufacturer, who is located at the end of the supply chain, will frequently receive only general information about the single‐use equipment components. Knowledge about the exact materials of construction is, however, crucial for suc­ cessful design of extractables studies. Ultimately, a com­ mon denominator for information exchange between the (bio)pharmaceutical manufacturer and the singleuse equipment supplier must be negotiated for the sake of patient safety. Single‐use equipment requires low acquisition costs but has relatively high arising expenses. Therefore, and because of the high qualification‐ and validation demand, single‐use equipment is usually procured from a single supplier. Double sourcing is rare. What is more, single‐ use equipment is often customized for the accordant application in terms of materials and dimensions. The resulting dependency of the (bio)pharmaceutical manu­ facturer on his single‐use equipment supplier in addition to the multilevel supply chain construct bares certain risks: delay in or stop of supply, mistakes due to falsely implemented or wrongly assembled plastic components that impede use of the single‐use equipment, unfavora­ ble and one‐sided price evolution, and change to plastic components that are made of plastic resins of lower qual­ ity. Change to another single‐use equipment supplier cannot be realized in short time. The single‐use equipment supplier can principally provide generic material qualification data namely, func­ tional, physicochemical, and biological parameters including chemical compatibility, extractables, and bio­ logical reactivity. Many suppliers do provide accordant data, however, reliability and consistency of the data should be verified by the (bio)pharmaceutical manufac­ turer. Validation guides and data sheets are often

11.5 Summary

marketing‐driven and provided without liability. The extractables data might not be appropriate for the intended application. Ultimately, it is the (bio)pharma­ ceutical manufacturer who is responsible for the suitabil­ ity of the used single‐use equipment that is used for production of the drug products. 11.4.2  Cost Factor Chemical safety assessment of single‐use equipment is cost‐intensive. An extractables study costs between €15 000 and €50 000 or more depending on the extent of the study (number of model solvents, analytical meth­ ods, etc.). For small plastic parts such as connectors, the costs for an extractables study might even exceed the purchasing costs. However, small plastic parts just as bigger parts, such as filters, can be the source of undesired leachables. Therefore, extractables testing is crucial. A typical single‐use equipment consisting of a filter, two types of tubing, two different connectors, a sensor, and a bag, thus seven different components and an aver­ age price of €30 000 per extractables study, would require some €210 000 for extractables testing only excluding material costs (samples). The costs for leachables studies are difficult to estimate. Depending on the target mole­ cule and the required validation extent, prices up to €50 000 per compound may occur. If the (bio)pharma­ ceutical manufacturing company has no in-house toxi­ cologists, additional costs for external toxicological assessment comparable to consulting costs must be cal­ culated. The (bio)pharmaceutical manufacturer can save costs if the single‐use equipment supplier provides extractables data of adequate quality. 11.4.3  Time Factor Chemical safety assessment of single‐use equipment and, in particular, generation of extractables data is time‐ intensive. In average, it takes three months to complete an extractables study (counting from an initial contact with the contract laboratory to the final report). If prob­ lems occur such as unidentifiable peaks in relevant con­ centrations, finding “the needle in the haystack” might consume ­additional weeks if not months. Subsequent simulated‐use extractables studies and/or leachables studies ­will prolong the chemical safety assessment sig­ nificantly. Toxicological assessment can be time‐con­ suming (weeks or even months) if no adequate toxicological data are available. For timely implementa­ tion of single‐use equipment, it is therefore important to consider the time requirement for chemical safety assess­ ment. To be on the safe side, the assessment should be started a year before the anticipated implementation date either clinical or commercial.

11.4.4  Production Outsourcing and Contract Manufacturers The (bio)pharmaceutical manufacturer may buy the API from other companies and/or outsource parts of the pro­ duction process to contract manufacturers. Marketing, however, obliges the (bio)pharmaceutical manufacturer to ensure chemical safety of the final drug product thus requiring adequate control procedures for the company/ contract manufacturer. This includes, among others, the choice of adequate plastic processing and packaging materials and generation of  meaningful extractables data. Any changes to the ­production process must be timely addressed to the (bio)pharmaceutical manufac­ turer. If the contract manufacturer uses own fabricated plastic materials, intellectual property might prevent the (bio)pharmaceutical manufacturer from investigating the material. In this case access to relevant data with regard to chemical safety of the material should be guar­ anteed in an accordant contract. 11.4.5  Life‐Cycle Management Due to the complex supply chain of single‐use equip­ ment and the number of involved suppliers and sub‐sup­ pliers, it is likely to receive change notifications referring to changes of suppliers, raw materials, and production sites. Any change in plastic resin or other raw material is a major change that can affect the functional, physico­ chemical, and biological parameters of the single‐use equipment, thus requiring update of the chemical safety assessment. Any change in a polymer is at risk to yield unfavorable extractables/leachables profiles, disqualify­ ing the single‐use equipment for its intended use. The (bio)pharmaceutical manufacturer should therefore agree with the single‐use equipment supplier to be timely notified, and if necessary, sufficiently long supplied with the current material until the change is thoroughly eval­ uated. Note that change of production to another coun­ try might have additional consequences such as change of raw materials suppliers. The (bio)pharmaceutical manufacturer is obliged to qualify his single‐use equip­ ment supplier(s). Initial qualification is followed by regu­ lar  re‐­qualification that is triggered by changes to the materials and/or the intended use. The frequency of the latter can increase in the case of reoccurring problems in terms of supply and quality.

11.5 ­Summary Single‐use equipment brings incomparable advantages in (bio)pharmaceutical manufacturing in terms of flexi­ bility and production time. One of the main disadvantages of single‐use equipment is the release of leachables into

155

156

11  Extractables/Leachables from Single‐Use Equipment

the drug product and therefore, risk for the efficacy and, in particular, the safety of the drug product and there­ fore, patient safety. Thus, chemical safety assessment of single‐use equipment is often crucial. The (bio)pharma­ ceutical manufacturer is responsible and liable for this chemical safety assessment. This chapter explains migration of extractables/leacha­ bles from plastic material into a contact medium such as a pharmaceutical drug product. It depicts the regulatory landscape and assigns responsibilities among the (bio) pharmaceutical manufacturer and the single‐use equip­ ment supplier and the sub‐suppliers. An approach to handle extractables/leachables from single‐use equip­ ment and the remaining production process is presented in detail offering a workflow for chemical safety assess­ ment, major aspects with regard to extractables studies, support in dealing with unknown extractables, and ways to perform toxicological risk assessment. Emphasis is put on the expenses for chemical safety assessment in terms of costs and time. Life‐cycle management and the derived tasks for the (bio)pharmaceutical manufacturer are acknowledged. The challenges that come with chemical safety assessment of single‐use equipment are addressed. Altogether, provided chemical safety and compliance to other functional, biological, and physicochemical requirements, single‐use equipment can be used with a clear conscience.

11.6 ­Discussion and Outlook Due to the number of advantages, plastic is, in general, increasingly used in (bio)pharmaceutical manufacturing with single‐use equipment setting a popular example. However, implementation, qualification, validation, and handling of plastic and, in particular, single‐use equipment represents a continuous challenge to the (bio)pharmaceuti­ cal manufacturer in many ways. Explicit regulatory require­ ments on single‐use equipment do not exist. On the one hand, the (bio)pharmaceutical manufacturer is liable for the safety of the plastic material used. On the other hand, suf­ ficient and adequate information about the different plastic materials is often not provided by the suppliers, mostly due to nondisclosure of proprietary information. The (bio) pharmaceutical manufacturer, however, usually does not have the expertise to take advantage of such proprietary information. The plastic manufacturer/supplier hence misses the chance to distinguish himself and to create com­ petitive advantage for further business relationship. A (bio)pharmaceutical company is not necessarily equipped with in‐house expertise about plastic materials, thus complicating qualification of single‐use equipment. In addition to the single‐use equipment supplier, installation of single‐use equipment might require a third‐party service provider, usually an engineering

company. This results in a three‐way collaboration between the (bio)pharmaceutical manufacturer, the sin­ gle‐use equipment supplier, and the engineering com­ pany, thus adding to the existing problems regarding information exchange and timelines. Single‐use equipment is vulnerable to airflow, heat, light (degradation of plastic), and mechanical damage due to the soft, flexible, and complex shape. In addition to safe installation, training of operators on handling of single‐ use equipment is of particular importance. Another aspect for the (bio)pharmaceutical manufacturer to con­ sider is shelf life of single‐use equipment along with the resulting in‐house supply and storage planning. In the EU, plastic material is usually provided with three years of shelf life, while US standard extends  to five years. Determination of shelf life falls to the plastic material supplier and it can differ between extensive physico­ chemical and biological testing to evaluation of parame­ ters according to Pharmacopeia testing. More than standardization of chemical safety assess­ ment including extractables profiling, there is a strong need to assign responsibility and liability for the safety of the polymer material to the accordant supplier and his sub‐suppliers. Given the interminable notifica­ tions about failed plastic material in pharmaceutical use due to discharge of too high quantities of leacha­ bles of toxicological concern, there is a need for produc­ tion of cleaner plastic materials and more transparency in the entire supply chain. It should become a legal responsibility of the supplier of single‐use equipment to specify the quality of their vendors’ incoming starter materials and thereby, to avoid unreasonably high amounts of impurities that could migrate into drug products and therefore, pose a risk for patients. Single‐use equipment and the individual plastic com­ ponents should be fabricated under controlled quality ­systems that grant quality key elements as laid down in ICH Q10. The (bio)pharmaceutical manufacturer is responsible, subject to good manufacturing practice and current regulations, to provide safe and effective drug products to the customers. The plastic manufac­ turer and supplier should support this with reliable, comprehensive, and exhaustive information and con­ sistent chemically safe products.

­Acknowledgments The authors would like to thank Dr. Andreas Nixdorf for the fruitful discussion of extractables and leachables from plastic materials and the consequences for use in (bio)pharmaceutical manufacturing in addition to the input on extractables profiling and review of this book chapter. Furthermore, the authors would like to thank the toxicologists Dr. Nina Macho and Dr. Silvio Wuschko for their significant contribution to the section on the toxicological risk assessment.

­  References

Nomenclature AET API BPOG CAS CFR EMA EP ESI EtO EU HPLC ICH ISO

Analytical evaluation threshold Active pharmaceutical ingredient BioPhorum Operations Group Chemical Abstracts Service Code of Federal Regulations European Medicines Agency European Pharmacopeia Electrospray ionization Ethylene oxide European Union High‐performance liquid chromatography International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use International Organization for Standardization

IR Infrared spectroscopy LOD Limit of detection LOQ Limit of quantification NaCl Sodium chloride NVR Nonvolatile residue PE Polyethylene PP Polypropylene PQRI Product Quality Research Institute PVDF Polyvinylidene fluoride REACH Registration, Evaluation, Authorisation and Restriction of Chemicals SCT Safety concern threshold TI Tolerable intake TTC Threshold of toxicological concern USP United States Pharmacopeia UV Ultraviolet

­References 1 Bee, J.S., Randolph, T.W., Carpenter, J.F. et al. (2011).

2

3

4

5

6

7

8

Effects of surfaces and Leachables on the stability of biopharmaceuticals. J. Pharm. Sci. 100 (10): 4158–4170. Liu, D., Nashed‐Samule, Y., Bondarenko, P.V. et al. (2012). Interactions between therapeutic proteins and acrylic acid leachable. PDA J. Pharm. Sci. Technol. 66 (1): 12–19. Li, K., Rogers, G., Nashed‐Samuel, Y. et al. (2015). Creating a holistic extractables and leachables (E&L) program for biotechnology products. PDA J. Pharm. Sci. Technol. 69 (5): 590–619. Gopalrathnam, G., Sharma, A.N., Dodd, S.W., and Huang, L. (2018). Impact of stainless steel exposure on the oxidation of Polysorbate 80 in Histidine placebo and active monoclonal antibody formulation. PDA Pharm. Sci. Technol. 72 (2): 163–175. Khanal, O., Singh, N., Traylor, S.J. et al. (2018). Contributions of depth filter components to protein adsorption in protein bioprocessing. Biotechnol. Bioeng. 115 (8): 1938–1948. Woodward, Z., Brooks, P., Morris‐Smith, B. et al. (2018). Adsorption and leachable contamination of Flucloxacillin, Cyclosporin and Amiodarone following delivery through an intravenous administration set. Pharm. Res. 35 (6): 121. SpecialChem. (2018). The universal selection source: polymer additives. http://polymer‐additives. specialchem.com. (accessed 2 August 2018). Peng, J., Zhao, Y., Hong, Y. et al. (2018). Chemical identity and mechanism of action and formation of a cell growth inhibitory compound from polycarbonate flasks. Anal. Chem. 90 (7): 4603–4610.

9 Jenke, D., Christiaens, P., and Iannone, J. (2017). PDA

10

11 12 13 14

15

16

17

18

19

extractables & leachables. Training Course (16–17 March). Barcelona. U.S. Food & Drug Administration. (2017). CFR Title 21: subchapter F‐biologics. Section 600.11 physical establishment, equipment, animals, and care. U.S. Pharmacopeia. (2016). USP containers – plastic. U.S. Pharmacopeia. (2017). USP plastic closures for injections. European Pharmacopeia. (2016). EP 3.1.3 plastics for pharmaceutical use – polyolefins. European Pharmacopeia. (2016). EP 3.1.4 plastics for pharmaceutical use – polyethylene without additives for containers for parenteral preparations and ophthalmic preparations. European Pharmacopeia. (2016). EP 3.1.5 plastics for pharmaceutical use – polyethylene with additives for containers for parenteral preparations and ophthalmic preparations. European Pharmacopeia. (2016). EP 3.1.6 plastics for pharmaceutical use – polypropylene for containers and closures for ophthalmic preparations. European Pharmacopeia. (2016). EP 3.1.15 plastics for pharmaceutical use – polyethylene terephthalate for containers not for parenteral use. European Pharmacopeia. (2015). EP 3.2.2 plastic containers and closures for pharmaceutical use. European Pharmacopeia. (2017). EP 3.2.9 rubber closures for containers and U.S. Food & Drug Administration. CFR Title 21. Subchapter F‐Biologics. Section 600.11.

157

158

11  Extractables/Leachables from Single‐Use Equipment

20 European Pharmacopeia. (2015). EP 3.1.9 silicone 21

22 23 24

25

26 27

28 29

30

31

32 33 34

35

36 37

38

39

40

elastomer for closures and tubing. International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use. (2014). ICH Q3D (step 4 version): guideline for elemental impurities. U.S. Pharmacopeia. (2017). USP elemental impurities – limits. U.S. Pharmacopeia. (2017). USP elemental impurities – procedures. U.S. Pharmacopeia. (2015). USP assessment of extractables associated with pharmaceutical packaging/ delivery systems. U.S. Pharmacopeia. (2015). USP assessment of drug product leachables associated with pharmaceutical packaging/delivery systems. U.S. Food & Drug Administration. (2017). 21 CFR 174 indirect food additives: general. U.S. Food & Drug Administration. (2017). 21 CFR 175 indirect food additives: adhesives and component coatings. U.S. Food & Drug Administration. (2017). 21 CFR 177 indirect food additives: polymers. U.S. Food & Drug Administration. (2017). 21 CFR 178 indirect food additives: adjuvants, production aids, and sanitizers. U.S. Food & Drug Administration. (2017). 21 CFR 179 irradiation in the production, processing and handling of food. U.S. Food & Drug Administration. (2017). 21 CFR 180 food additives permitted in food or in contact with food on an interim basis pending additional study. U.S. Food & Drug Administration. (2017). 21 CFR 181 prior‐sanctioned food ingredients. U.S. Food & Drug Administration. (2017). 21 CFR 182 substances generally recognized as safe. U.S. Food & Drug Administration. (2017). 21 CFR 184 direct food substances affirmed as generally recognized as safe. U.S. Food & Drug Administration. (2017). 21 CFR 186 indirect food substances affirmed as generally recognized as safe. U.S. Food & Drug Administration. (2017). 21 CFR 189 substances prohibited from use in human food. Karamfilova, E. (2016). Food Contact Materials – Regulation (EC) 1935/2004 European Implementation Assessment. European Medicines Agency. (2005). Guideline on plastic immediate packaging materials. CPMP/ QWP/4359/03 and EMEA/CVMP/205/04. Stults, C.L.M., Ansell, J.M., Shaw, A.J., and Nagao, L.M. (2015). Evaluation of Extractables in processed and unprocessed polymer materials used for pharmaceutical applications. AAPS PharmSciTech. 16 (1): 150–164. McKeen, L.W. (ed.) (2018). The Effect of Sterilization on Plastics and Elastomers, 4e. Oxford: Elsevier Inc.

41 Magarian, N., Lee, K., Nagpal, K. et al. (2016).

42

43

44

45

46

47

48

49

50

51

52

53

54

Clearance of extractables and leachables from single‐ use technologies via ultrafiltration/diafiltration operations. Biotechnol. Prog. 32 (3): 718–724. U.S. Pharmacopeia. (2017). USP draft: polymeric components and systems used in the manufacturing of pharmaceutical and biopharmaceutical drug products. U.S. Pharmacopeia. (2017). USP draft: plastic components and systems used to manufacture pharmaceutical drug products. Nouman, M., Saunier, J., Jubeli, E., and Yagoubi, N. (2017). Additive blooming in polymer materials: consequences in the pharmaceutical and medical field. Polym. Degrad. Stab. 143: 239–252. Paskiet, D., Jenke, D., Ball, D. et al. (2013). The product quality research institute (PQRI) leachables and extractables working group initiatives for parenteral and ophthalmic drug product (PODP). PDA J. Pharm. Sci. Technol. 67 (5): 430–447. Ding, W., Madsen, G., Mahajan, E. et al. (2014). Standardized Extractables testing protocol for single‐use systems in biomanufacturing. Pharm. Eng. 34 (6): 1–11. Kim, J.J. and Lee, K.T. (2012). Determination of monomers and oligomers in polyethylene terephthalate trays and bottles for food use by using high performance liquid chromatography‐electrospray ionization‐mass spectrometry. Polym. Test. 31 (3): 490–499. International Standard. (2015). ISO 10 993 Part 17. Biological evaluation of medical devices – establishment of allowable limits for leachable substances. Sobańtka, A, Brandstätter, H, and Weiner, C. (2016). The fallacy of extractables determination via non‐ volatile residue (NVR). E&L Europe (08–10 November). Dublin. International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use. (2017). ICH M7 (R1): assessment and control of DNA reactive (Mutagenic) impurities in pharmaceuticals to limit potential carcinogenic risk. Lapenna, S. and Worth, A. (2011). Analysis of the cramer classification scheme for oral systemic toxicity – implications for its implementation in Toxtree. EUR 24898 EN. ISBN 978‐92‐79‐20804‐1. doi: https://doi.org/10.2788/39716. International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use. (2011). ICH Topic Q3C (R5): impurities: guideline for residual solvents. European Medicines Agency. (2008). Guideline on the specification limit for resuidues of metal catalysts or metal reagents. EMEA/CHMP/SWP/4446/2000. Jenke, D. (2017). Extractables screening of polypropylene resins used in pharmaceutical packaging for safety hazards. PDA J. Pharm. Sci. Technol. 71 (5): 346–367.

159

12 The Single‐Use Standardization P.E. James Dean Vogel The BioProcess Institute, North Kingstown, RI, USA

12.1 ­Introduction Within the last decade, the single‐use technology (SUT) arena of the biopharmaceutical industry has exponen­ tially increased. Leading organizations have adopted the ground‐breaking advantages of these components over traditional multiuse parts and technologies to deliver safer drugs with less risk. The initial value and risk‐ reduction results are being realized, but not without the emergence of other trade‐offs. More and more end users are calling for standardization in emerging areas of bio­ pharma, while also recognizing the challenges posed by trying to stay current with these standards. How do we, as an industry, accomplish this? Regulatory agencies encourage the use of consensus for standardiza­ tion, and single‐use component suppliers and end users have begun the process of harmonizing, collaborating, and standardizing. This is the perfect opportunity for the best and brightest in our industry to collaborate their time and knowledge on this process. The organizations that write the standards or guidelines for this community are comprised of our peers, qualified individuals from these very same companies. We as an industry need to do a better job of capturing what the current state is. What is possible? And what is needed to better the pro­ cess? This chapter provides an overview on the current state of single‐use standardization.

12.2 ­Alphabet Soup Many leaders in the biopharmaceutical industry, known by their acronyms, have been involved with SUT. As SUT has become less of a novelty and more of a necessity in the pur­ suit for industry standardization, organizations such as American Society of Mechanical Engineers‐BioProcessing Equipment (ASME‐BPE), ASTM International (formerly named American Society for Testing and Materials,

ASTM), BioPhorum Operations Group (BPOG), Bio‐ Process Systems Alliance (BPSA), Society for Chemical Engineering and Biotechnology (DECHEMA Gesellschaft für Chemische Technik und Biotechnologie e.V.), The Extractables and Leachables Safety Information Exchange (ELSIE), International Society for Pharmaceutical Engineering (ISPE), Parenteral Drug Association (PDA), Product Quality Research Institute (PQRI), and United States Pharmacopeia (USP) have all been involved (see  Table  12.1). In the time since the SUT collaborative community has been including these groups, some posi­ tive signs of progress have slowly surfaced. Challenges, however, still exist. How do we decide when novel becomes standard? Who decides it? Most people are calling for standardization of SUT, and some have tried, but it simply has not happened yet. Why not? It could, in part, be that standardization means differ­ ent things to different people. To some, standardization simply means ensuring connectivity between compo­ nents and uniform vessel dimensions. To others, it means a consensus body such as the ASME or ASTM International publishing an industry consensus method or specification that people will use and conform in order to develop the best and safest therapeutics. To most others, standardization, at the very least, means a minimum requirement and/or general guidance based upon best practices and/or the current state‐of‐the‐ industry. This is a new and rapidly changing industry, which makes standardization complicated in its ability to stay current. Conformance may be perceived as a threat to growth and innovation. Perception becomes reality to some. At a minimum, the industry is in des­ perate need of a standard set of single‐use terms, which people can use to speak the same language. Seven years ago, a paper focusing on model solvents entitled “Standardization of Single‐Use Components Extractables Studies” was written by the members of ISPE’s Disposables Community of Practice [1]. The

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

Table 12.1  Brief descriptions of the organizations (in alphabetical order) currently at the single‐use community. Organizations

Description

Web

ASME‐BPE

Established over 130 years ago, the American Society of Mechanical Engineers is a group comprised of individual members from engineering firms, suppliers, consultants, and end‐user companies who focus on the design, education, and specifications of engineered equipment. ASME formed its Bioprocessing Equipment (BPE) Committee in 1990 to satisfy the need for standards in the design and cleanability of bioprocessing equipment. As their website states, the ASME‐BPE Standard is intended for design, materials, construction, inspection, and testing of vessels, piping and related accessories such as pumps, valves, and fittings for use in the biopharmaceutical industry. This American National Standards Institute (ANSI)‐approved standard provides requirements for systems and components that are subject to cleaning and sanitization, sterilization, and/or other suitable processes used in the manufacturing of biopharmaceuticals. This standard also provides requirements for single‐use systems and components used in the above listed systems.

www.asme.org

ASTM International

ASTM was established in 1898 to develop voluntary consensus standards for materials, products, systems, and services. ASTM E55 was formed in 2003 with the objective of developing international standardized nomenclature and definition of terms, recommended practices, guides, test methods, specifications, and performance standards for the manufacture of pharmaceutical products. In 2006, their scope was expanded to address all aspects of biopharma including the manufacture of pharmaceutical products. All members have a say through consensus‐based decision making. These members come from manufacturers, regulatory agencies, associations, consultancies, academia, labs, and research institutes. ASTM International now includes 144 technical committees and offices worldwide. Today, 12 396 ASTM standards are used around the world to improve product quality, enhance safety, facilitate market access and trade, and build consumer confidence. ASTM International includes more than 35 000 of the world’s top technical experts and business professionals, representing 150 countries. Working in an open and transparent process and using ASTM’s advanced electronic infrastructure, ASTM members deliver the test methods, specifications, guides, and practices which support industries and governments worldwide (http://www.astm. org/ABOUT/overview.html)

www.astm.org

BPOG

BPOG is a company‐to‐company collaboration for the biopharmaceutical manufacturers or end users. Member companies agree upon their strategic priorities and then nominate their top subject‐matter experts to share and agree best practices. In all, there are 23 member companies fielding over 700 representatives in 12 work streams. BPOG has established best practices on a wide range of quality, engineering, and organizational topics central to the challenge of mastering biotech drug substance operations.

www.biophorum.com

BPSA

The BPSA is an industry‐led corporate member industry‐led trade association dedicated to encouraging and accelerating the adoption of single‐use manufacturing technologies used in the production of biopharmaceuticals and vaccines. Its membership consists of 45 companies and is made up of industry suppliers with 10+% end users. BPSA facilitates education, sharing of best practices, development of consensus guides, and business‐to‐business networking opportunities among its member company employees. This is the only organization solely focused on SUT, and they hold their Annual International Single‐Use Summit in Washington, D.C. every July.

www.bpsalliance.org

DECHEMA

DECHEMA e.V. is the expert network for chemical engineering and biotechnology in Germany. It is a nonprofit professional society representing these fields in science, industry, politics, and society. Around 100 subject divisions and working groups offer a forum for knowledge exchange among experts based on trust. DECHEMA consolidates the know‐how of more than 5800 individual and sustaining members. The DECHEMA working group “Single‐Use Technology in Biopharmaceutical Manufacturing” was founded in 2010. It currently consists of 150 members including both suppliers and users, mainly from German‐speaking countries, and meets two times per year. The working group develops practice‐oriented recommendations on a number of topics and thus supports the increasing implementation of single‐use systems. Among the current topics of the working group are process intensification and continuous bioprocessing, extractables and leachables detection, as well as sustainability of single‐use technology.

www.dechema.de

12.4  Compare and Contrast

Table 12.1  (Continued) Organizations

Description

Web

PDA

PDA is the worldwide leading provider of science, technology, and regulatory information and education for the pharmaceutical and biopharmaceutical industries. Founded in 1946 as a nonprofit organization, PDA now has over 9500 members worldwide. Using their expertise, these members are committed to developing scientifically sound technical information for practical uses in order to advance science and its regulations.

www.pda.org

PQRI

PQRI is a nonprofit group of organizations who collaborate to create important information that improves drug product quality and development. This information is shared among the participating organizations. PQRI, through its varied collaborations, provides a distinctive environment; focusing input to create a broad and timely impact on regulations, methodology, and creation of industry standards.

www.pqri.org

USP

The USP is a scientific nonprofit organization that sets standards for the quality, purity, strength, and identity of medicines, food ingredients, and supplements. USP’s drug standards are enforceable in the United States by the FDA. These standards are also used in more than 140 other countries.

www.usp.org

These organizations have many things in common and all are gold mines full of information, innovation, and know‐how.

authors referenced the BPSA’s definition of extractables and leachables and proposed some common solutions in the biopharmaceutical industry; these solutions which, if exposed to the SUT, should help compare dif­ ferent SUTs and their potential risk to the process solu­ tion and drug product. This set off a debate that is still alive years later. The focus of this debate is whether these model solvents generate extractables, or leach­ ables, or something in between. BPOG has embraced the spirit of the paper and published their own formal approach to these model solvents in 2014 [2]. Other organizations also appreciate the concept, while disa­ greeing with some specifics; openly acknowledging that we, the SUT community, can do better. Through recent organizational efforts and collaborations, we have dis­ cussed the need to formally identify these model sol­ vents and classify them within existing and well‐established definitions utilized by USP and PQRI. The complexity of single‐use components and the number of participants can complicate the process of collaboration on the path to producing better and safer drugs. Future develop­ ments and implementations of new SUTs in the bio­ pharmaceutical industry would also significantly benefit from the development of this shared information and a willingness to communicate with one another for the good of the whole.

present similar risks, if not more risks to both materials and supply chain. Food regulations provide a good start, but are not intended for managing biopharmaceuticals. The SUS industry has also borrowed from the medical device industry some guidance such as USP Class VI and International Organization for Standardization (ISO) 10993 for biocompatibility (which have been ref­ erenced for years) and ISO 11137 (sterilization of healthcare products). Companies have also applied drug product standards, final vial and stopper requirements have been referenced for years, and drug product requirements such as USP 788 for particulates have been referenced. Filters comply with the 21 Code of Federal Regulations 211.72 requirement as to not to release fibers into products. As with everything, the challenge is that we have more information these days, and the requirements borrowed from other industries are becoming less applicable. Standards and codes are developed and written pre­ sumably to allow their content to withstand the rigors of jurisprudence, should the engineering integrity of their use be questioned in a court of law. A guideline, on the other hand, serves to bring together best practices from a broad range of sources, regarding system design, vali­ dation, and commissioning. This is vital to keep in mind as we move forward standardizing as an industry.

12.3 ­History

12.4 ­Compare and Contrast

Food industry guidelines have historically been the requirement for seals, gaskets, and hoses used with tra­ ditional equipment. Although they have been over­ looked and surpassed by single‐use systems (SUSs), they

We have seen with these efforts that the key is to under­ stand one another. It is important to note some differ­ ences between the groups and to appreciate each organization and how they work. Membership in each

161

162

12  The Single‐Use Standardization

can be classified into one of two categories: corporate (e.g. BPSA, BPOG, and ELSIE) vs. individual (e.g. ASME‐ BPE, ASTM, DECHEMA, ISPE, PDA, and PQRI). Of course, the people within individual member organiza­ tions represent their company, but it is not as formal as the corporate member organizations in which their view may be more coordinated and valued. The fact that the charters of these member organizations are structured in different ways can have a positive effect on collaboration, varying vantage points can often lead to a more accurate picture of the whole. As SUT use rises, it evolves. The communities form their best practices and guidelines. Then, the organiza­ tions formalize them into technical reports and other documents to enhance those guidelines and bring them to the next level of convention. The next stage is when the consensus standards build off of those. As use increases, requirements are added, and then used for ref­ erence by regulators and users. Figure 12.1 shows the levels and relationships between the “lettered” organizations. There is interaction among all levels of this diagram. But executing the interactions correctly in order to balance innovation and the develop­ ment of the technology is the trick. ●●

●●

●●

●●

At the foundation, BPSA and BPOG – industry con­ sortia – are corporate members representing corpora­ tions, end users, and suppliers. They have the flexibility to present best practices and publish documents. Next are PDA and ISPE who are industry organiza­ tions comprised of individuals. They publish best practice documents and technical reports that are routinely referenced by standards and compendia organizations as best practices. DECHEMA is a non­ profit society with members from industry and aca­ demia and publishes recommendations for the SUT community. ASTM and ASME are recognized ANSI consensus standards organizations. The characteristics of an ANSI standard include the following: consensus by a group that is open to representatives from all inter­ ested parties; broad‐based public review and comment on draft standards; considerations of, and response to, comments; incorporation of submitted changes that meet the same consensus requirements into a draft standard; and availability of an appeal by any partici­ pant who alleges that these principles were not respected during the standards‐development process. The Food and Drug Administration (FDA) encourages use of these standards where appropriate. USP and European Pharmacopeia (EP) are compendia organizations which develop standard legally binding testing methods required by law. The U.S. Pharmacopeial Convention is a private, nongovernmental organization

●●

that publishes the USP and the National Formulary (NF) as official compendia of the United States. Although much of the USP and NF is legally enforce­ able, the USP general chapters numbered above (general information chapters) are informational and generally do not contain any mandatory requirements. General information chapters might include some rec­ ommendations that may help a firm meet. In contrast, the European Pharmacopeia Commission is the govern­ ing body of the EP, responsible for overseeing the practi­ cal work of experts in every field of the pharmaceutical sciences – all volunteers – who participate in groups of experts and working parties. The European Medicines Agency (EMA) works with the European Directorate for the Quality of Medicines and HealthCare (EDQM), a directorate of the Council of Europe. The EDQM is an organization that protects public health by enabling the development, supporting the implementation, and monitoring the application of quality standards for medicines and their safe use. The EP provides a legal and scientific reference for the quality control of medi­ cines. All producers of medicines and substances for pharmaceutical use must apply these quality standards in order to market their products in the signatory states of the Convention. This means that companies must follow these standards when applying to the EMA and include reference to the monographs in the quality part of their applications. Last, but not least, the FDA  –  and other regulatory agencies – establish law and create regulations for the industry. As noted earlier, in 1996, the U.S. Congress passed the National Technology Transfer and Advancement Act (NTTAA) (PL104‐113) which, according to the FDA’s website, directs “agencies to use voluntary consensus standards in lieu of government‐ unique standards except where inconsistent with law or otherwise impractical.” As part of NTTAA, the FDA is required to prepare yearly reports on the use of standards.

As Figure 12.1 illustrates, as you move upward through this list, flexibility decreases while regulatory power increases. Conversely, as you move downward, regula­ tory authority lessens and flexibility broadens.

12.5 ­Collaboration and Alignment Lead to Standardization The industry is evolving and progress is being made. The organizations now call each other seeking guidance, or to provide formal correspondence and updates of the other group’s SUT activities. There are the beginnings of formal cross‐organization reviews of draft documents. It

12.6  General SUT Efforts

Figure 12.1  Regulation vs. flexibility.

is all heading in the right direction, but more is needed in this important and rapidly changing field. We must adapt and overcome the challenges together in order to reach standardization.

Box 12.1  Sections Included in the Single‐use Appendix ●● ●●

12.6 ­General SUT Efforts

●● ●● ●●

ASME‐BPE has been very active and formed many SUT‐related task groups to address the issues of the industry. The work started in the 1990s with some of the first industry guidance on polymeric materials. SuSs and components were added to the scope of the BPE in 2014, and a mandatory appendix specifically for single‐use will be in its 2019 edition (see Box 12.1). The BPE is continuing work with active task groups in integrity, particulates, and connectors. The BPE also added a section on change management on part project management in 2019. This built off the BPSA/BPOG change notification paper and lists four types of changes and responsibilities for the owner/user and the manufac­ turer. ASME has also aligned its definition of “Product & Process Contact Surface” with ISPE’s new “Baseline Guide” in the 2014 edition. BPSA published the “Role of Single‐Use Polymeric Solutions in Enabling Cell and Gene Therapy Production” in 2018. This provides guidance for the emerging cell and gene therapy area, where particulate and integrity requirements are higher than traditional single‐use.

●●

●● ●● ●● ●● ●● ●● ●●

General guidelines for non‐CIP/SIP systems Polymeric materials BioCompatability Extractables and leachables Components (polymeric hygienic unions, steam through/thru) Polymeric joining (barbs and thermoplastic elastomer welding) Identification and labeling Certificate of compliance Inspection and packaging Sterilization Shelf life, storage, and expiration Particulates Methods required by law.

BPSA also published the “Consensus Quality Agreement for Single‐Use Biopharmaceutical Manufacturing Products” which is intended to give an example of com­ mon structure and to facilitate a dialog between suppli­ ers and end users for quality agreements. Such a dialog will reveal the specific needs of the end user and the many attributes of the quality system and operating mechanisms of the supplier, which in turn will provide the content for the final quality agreement between the

163

164

12  The Single‐Use Standardization

two parties. In addition, a revision of “Component Quality Test Reference Matrices” is now being updated by BPSA to include the addition of new sections as well as current topics such as filters, tubing, connectors, and bags. DECHEMA: The DECHEMA working group “Single‐ Use Technology in Biopharmaceutical Manufacturing” has published the online encyclopedia “Single‐Use Technology from A to Z” and the status paper “Facility of the Future.” These two publications are directed at first‐ time users from both academia and industry. In addition, the DECHEMA recommendations “Recommendation for a Risk Analysis for Production Processes with Disposable Bioreactors” and “Recommendations for Process Engineering Characterization of Single‐Use Bioreactors and Mixing Systems by Using Experimental Methods” support the selection, implementation, and comparison of single‐use and reusable devices. The authors of all DECHEMA documents include both sup­ pliers and end users. ISPE’s Disposables Community of Practice group pub­ lished a comprehensive “Good Practice Guide for Single‐ Use Technology” in 2018. The guide is a practical go‐to document providing a road map for implementing sin­ gle‐use products. The authors of this document include end users and suppliers. PDA’s “Technical Report No. 66: Application of Single‐ Use‐Systems in Pharmaceutical Manufacturing” was published in the fall of 2014. PDA Scientific Advisory Boards, PDA members, and PDA staff are now in the process of discussing next steps in supporting the imple­ mentation of SUT. PDA hosted a workshop at PDA headquarters in Bethesda, MD, in 2014, entitled “Single‐ Use Systems Cross‐Organizational Workshop.” Twenty industry representatives and ten FDA staff attended the meeting. Representatives from ASME, ASTM, BPOG, BPSA, ELSIE, PDA, PQRI, and USP attended and made formal presentations. The objective of this workshop was twofold: (i) to promote a harmonized approach to supporting SUS activities within the industry and, in so doing, to minimize duplication of efforts and (ii) to com­ municate ongoing SUS initiatives among the group. PDA is committed to hosting the meeting again.

12.7 ­Leachables and Extractables Members of the alphabet soup have discussed the need to separate the terms leachables and extractables, because they are not always addressed in tandem (see also Chapters 8, 11, and 17–19). BPOG published in 2017 the “Best Practices Guide for Evaluating Leachables Risk from Polymeric Single‐Use Systems Used in Biopharmaceutical Manufacturing” to help define the

differences between extractables and leachables from many perspectives. USP revised Chapter , specifically to address plastic packaging systems and their plastic materials of construction. Specific chapters include: ●●

●● ●●

●●

Plastic Packaging Systems and Their Materials of Construction Plastic Materials of Construction Plastic Packaging Systems for Pharmaceutical Use and Evaluation of Plastic Packaging Systems and Their Materials of Construction With Respect to Their User Safety Impact.

They are published in 2018 and are targeted to be offi­ cial in 2020. Recognizing both the similarities and differences between packaging systems and systems used in the manufacturing of pharmaceuticals, USP is also develop­ ing a Chapter Polymeric Components and Systems Used in the Manufacturing of Drug Products to address plastic systems used in manufacturing and their materi­ als of construction. As SUT is a plastic system used in manufacturing, it is reasonable to expect that SUT would be covered by this section. An expert panel, which included members from BPSA, BPOG, PQRI, and ELSIE as well as experts from individual companies who are active in pharmaceutical manufacturing, was seated to develop this section. The DECHEMA working group “Single‐Use‐Technology in Biopharmaceutical Manufacturing” developed a “Recommendation for Leachables Studies – Standardized cell culture test for the early identification of critical films” in 2014. It enables the early identification of non‐satisfac­ tory films for the cultivation of Chinese hamster ovary (CHO) cell lines in chemically defined culture media. This test is based on a commercially available CHO cell line and has already been implemented both by suppliers and users in Europe. ELSIE continues to populate the Safety Information Database with safety information extractables toxicol­ ogy. Additionally, ELSIE has prepared three papers on evaluation of results from a controlled extraction study Pilot Program.

12.8 ­Particulates in SUT BPSA issued “Recommendations for Testing, Evaluation, and Control of Particulates in Single‐Use Process Equipment” in 2014, which was developed by subject‐ matter experts from both the supplier and end‐user sides of the bioprocess industry  –  in order to meet a need. This informative document not only helps readers to

12.13  SUT Design Verification

characterize and quantify types and levels of particles in SUSs and components but it also recommends proce­ dures for minimizing these types and levels. In addition, the document offers a look ahead to the future and what the industry should expect by way of improvements in particle control in order to protect the health and safety of pharmaceutical patients. The ASME‐BPE referenced this work in their section and there is still an industry need for collaboration on the methods to detect particles in the SUT fluid path, for which BPE and ASTM have formed a task group. ASTM’s work group WK43742 is writing a draft stan­ dard entitled “New Practice for Characterizing Particulates Burden” from SuSs. USP has issued acceptance limits for visible and sub­ visible particles. These can be found in Subvisible Particulate Matter in Therapeutic Protein Injections; Particulate Matter in Injections; Visible Particulates in Injection; and Visual Inspection of Injections (see also Chapter 18).

12.9 ­Change Notification BPOG and BPSA shared the same position about the need to standardize supplier‐change notifications. They worked together to publish “An Industry Proposal for Change Notification Practices for Single‐Use Biomanufacturing Systems” in 2017. They defined levels of change and to provide a standardized template for suppliers to streamline the change‐notification process and facilitate the implementation process for users (see also Chapter  18). This was used by the ASME‐BPE to generate their change management section. BPOG published a separate paper “A guide to the clas­ sification of changes to single use biomanufacturing sys­ tems” in 2018 to help implement these practices. ASTM is currently working on development of a change‐notification standard similar to “E2500 Standard Guide for Specification, Design, and Verification of Pharmaceutical and Biopharmaceutical Manufacturing Systems and Equipment.”

12.10 ­SUT System Integrity BPSA published the “Design, Control, and Monitoring of Single‐Use Systems for Integrity Assurance” in 2018. It describes key items to consider when evaluating SUT integrity. ASME‐BPE has formed a task group to address on integrity and ASTM formed a work group WK43741 “New Practice for Testing Integrity of Single‐Use Systems” which provides a standardized approach for SUT vendors

to use throughout the life cycle of the SUT component. It is intended to help the supplier measure an SUT system’s integrity based on risk assessment and can be applied to all categories of SUT components. USP Sterile Product – Package Evaluation pro­ vides an overview of “Leak Test” methodologies (also termed technologies, approaches, or methods) as well as “Package Seal Quality Tests” useful for verification of ster­ ile product package integrity. More detailed recommenda­ tions for the selection, qualification, and use of leak test methods are presented in three subchapters that address these specific topics: (i) Package Integrity and Test Method Selection; (ii) Package Integrity Leak Test Technologies; and (iii) Package Seal Quality Test Methods. At this time, there have been an increasing number of discussions about how to apply pres­ sure/rating to SUT (see also Chapters 2, 8, and 18).

12.11 ­SUT User Requirements BPSA has published the “Single Use User Requirements ToolKit Pack” to help specify SUT on the topic of supply chains in SUT. They continue to conduct webinars and write papers on this topic as well.

12.12 ­Connectors This is an area where there has been much discussion. Novel aseptic connectors have been developed from a hand full of suppliers (see also Chapter 5). Their develop­ ment is key for the ability to make aseptic connections with SUT and the adaption of the technology. Each connector is a unique design and offers its own advantages, e.g. fluid flow, ease of use. But, they are not compatible with each other. The community is calling for some type of “connec­ tivity” or is it “standardization”? The discussions have started to surface in the organizations. ASME‐BPE formed the Polymeric Hygienic Unions Task Group which is simi­ lar to their Metal Fitting Task Group, and a group on hose barbs. These discussions will continue, but most likely the market will drive any form of standardization, as with the  Universal Serial Bus for computers. The community needs to be cautious so not to let this become a Video Home System vs. Betamax situation and where are they now? Verses from the newer technology such as Digital Video Recorder or streaming video?

12.13 ­SUT Design Verification ASTM had published its standard E3051 “Standard Guide for Specification, Design, Verification, and Application of  Single‐Use Systems in Pharmaceutical and

165

166

12  The Single‐Use Standardization

Biopharmaceutical Manufacturing.” It is complimentary to the very successful standard for traditional equipment E2500. It is intended to be used by suppliers, sub‐suppli­ ers, and end users of SUT and takes into consideration the unique traits of single‐use components and their rising popularity in the biopharmaceutical industry.

12.14 ­Summary and Conclusions We need to be cautiously optimistic about where we are going, with and without standardization. This is basic human nature and up to all of us in the industry. To quote Einstein, “We can’’t solve problems by using the same kind of thinking we used when we created them.”

We are starting to make progress, but it is imperative that we, as an industry, set a course for standardization. Each of these lettered groups is made up of industry professionals and there is always room for more. Everyone is encouraged to research these organizations and find a place where they might add value to this effort. There is plenty of room for everyone. Bioprocessing experts are highly sought after in the industry for their experience, opinions, and guidance. Those with less expertise are also encouraged to par­ take of these volunteer opportunities in order to gain more knowledge and to contribute to shaping the future of the industry. You are the qualified people who can write and proof these standards.

Nomenclature ANSI American National Standards Institute ASME‐BPE American Society of Mechanical Engineers‐BioProcessing Equipment Standard ASTM American Society for Testing and Materials BPE BioProcessing Equipment BPOG BioPhorum Operations Group BPSA Bio‐Process Systems Alliance CHO Chinese hamster ovary DECHEMA Gesellschaft für Chemische Technik und Biotechnologie e.V. (Society for Chemical Engineering and Biotechnology) EDQM European Directorate for the Quality of Medicines and Health Care ELSIE The Extractables and Leachables Safety Information Exchange

EMA European Medicines Agency EP European Pharmacopeia FDA Food and Drug Administration ISO International Organization for Standardization ISPE International Society for Pharmaceutical Engineering NF National Formulary NTTAA National Technology Transfer and Advancement Act PDA Parenteral Drug Association PQRI Product Quality Research Institute SUS Single‐use system SUT Single‐use technology USP United States Pharmacopeia

References 1 Mahajan, E., Ray‐Chaudhuri, T., and Vogel, J.D. (2012).

Standardization of single use component’s extractable studies for industry. Pharm. Eng. 32 (3): 1–3.

2 Ding, W., Madsen, G., Mahajan, E. et al. (2014).

Standardized extractables testing protocol for single‐use systems in biomanufacturing. Pharm. Eng. 34 (6): 1–11.

Further Reading Articles Martin, J., Hartzel, W., Vogel, J. et al. (2009). Organizing the organizations of single‐ use manufacturing. BioProcess Int. 7 (Suppl. 4): 9–12. Pauley, J. (2013). Do businesses really need standards? Standards as strategic business tools. ASTM Standardization News December: 34–35.

Rios, M. (2014). Efforts toward the harmonization of single‐use standards. BioProcess Int. 10 (4): 4–11. Vogel, J.D. (2012). The maturation of single‐use applications: the state of the industry. BioProcess Int. 8 (Suppl. 5): 10–18. Whitford, W. and Galliher, P. (2014). Trends in setting single‐ use technology standards. BioPharm Int. 27 (2): 16–18.

References

Website References http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/ cfstandards/search.cfm www.asme.org www.astm.org www.bpsalliance.org www.dechema.de www.elsiedata.org http://www.fda.gov/MedicalDevices/DeviceRegulationand Guidance/GuidanceDocuments/ucm077274.htm

http://www.fda.gov/MedicalDevices/ DeviceRegulationandGuidance/Standards/default.htm www.ispe.org www.pda.org www.pqri.org www.usp.org

Town Hall References 2007 ISPE’s Annual Meeting, Las Vegas, NV (4–7 November). BPSA International Single‐Use Summits (July 2011–2014). Town Hall Forum: Single‐Use Systems, IBC Life Sciences, Boston, MA (22 October 2013). Town Hall Forum: Harmonization of Single‐Use Systems, BioProcess International Theater Round Table, INTERPHEX, New York, NY (17 March 2014).

Town Hall Forum: Change Control and Change Management in Single‐Use Systems, IBC Life Sciences BDP Week, San Diego, CA (26 March 2014). Town Hall Forum: Harmonization of Single‐Use Systems: Particulates; IBC Life Sciences 11th Annual Single‐Use Applications for Biopharmaceutical Manufacturing Meeting, Boston, MA (10 June 2014). Town Hall Forum: Standardization of Single‐Use Systems; BioProcess International Conference & Exhibition, Boston, MA (23 October 2014).

167

169

13 Environmental Impacts of Single‐Use Systems William G. Whitford1, Mark A. Petrich2, and William P. Flanagan3 1

GE Healthcare, South Logan, UT, USA Merck & Co., Inc., West Point, PA, USA 3 Aspire Sustainability, Albany, NY, USA 2

13.1 ­Introduction

13.2 ­Sustainability

Single‐use (SU) technology has become an important part of biotechnology research and commercial biomanufacturing. Consumption is increasing and so are concerns about the possible environmental impact of use and disposal of SU materials. SU technology can appear to be environmentally unfavorable compared with ­traditional biomanufacturing options. However, this is a perception driven by the visible quantities of packaging materials and spent SU components. Scientific evaluations using Life Cycle Assessment (LCA) show that SU technologies can be less environmentally impactful than the alternatives. Materials handling and disposal are important operational considerations, but disposal of used materials has a small contribution to the overall sustainability picture. Suppliers and end users should make use of scientific evaluation methods as they otherwise continue to reduce the environmental impact of SU technology. There are many types of “pollution” to be concerned about. A factor in debating green issues is that there are different forms and sources of pollution that affect different geographical locations to different degrees. ­ Furthermore, there are often very different approaches to solving a pollution problem, ranging from prohibiting the source of the pollution, blocking its detrimental effects, or establishing a systematic cleanup of the damage caused. There can often be “trade‐offs” involving offsetting of a negative environmental impact by a positive effect, often in a distinct and unrelated impact category. Finally, there can be honest disagreement as to both the types and extent of damage being caused and of the potential unintended collateral consequences of action taken to alleviate the problem [1].

Sustainability, as considered here, is a long‐term holistic approach that evaluates how biological systems remain diverse and productive over time, and also ­considers concerns of a more immediate and nonbiological focus [2]. Examples of this last point might be  the esthetics of a town’s solid waste disposal or a  nation’s  mineral resource depletion. The total concept  of sustainability has developed to often include three “pillars”: 1) Cultural preservation and social equity. 2) Economic development and technological progress. 3) Environmental consciousness and resource conservation. For the most part in this chapter, use of the term “sustainability” will emphasize the ability to use natural resources in a way that indefinitely protects the integrity (and limits the fouling or depletion) of existing biological and other environmental systems. Table  13.1 defines important terms generally employed in this field and that will be used throughout this chapter.

13.3 ­The Evolution of SU Technologies Traditional biologics manufacturing approaches involve the use of durable bioreactors and fluid containers in processes that consume large quantities of water and energy and require the use of detergents, caustics, and other materials that can be environmentally deleterious [3, 4]. SU approaches were initially thought to appear rather environmentally unfriendly due to the need for disposal of SU components. Meanwhile, sustainable manufacturing practices have been gaining favor for

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

170

13  Environmental Impacts of Single‐Use Systems

Table 13.1  Summary of terms having importance for this chapter. Term

Definition

Biomanufacturing

The manufacturing of cells, viruses, proteins, antibodies, hormones, growth factors, and nucleic acids for use as bio‐ and cell therapeutics. Biomanufacturing also includes waste processing using microbial systems. Synonyms include bioprocessing and bioproduction.

CIP/SIP

Clean‐in‐place and steam‐in‐place are cleaning and steam‐sterilizing durable equipment, usually in its installed location, between uses.

Consumables

Process buffers, solvents, culture media and resins, and SU components.

Coproducts

A second product intended for commercial use and is ordinarily used in the form produced by the process.

Durable process

A series of manufacturing operations composed of permanent equipment made of cleanable and reusable materials that are used repeatedly with no major components that are used only once. Synonyms include traditional, stainless steel, glass and steel, reusable, and classical.

End user

Either the company employing bioprocessing to manufacture a product or the customer employing the final biological product (in this chapter, it usually refers to the biomanufacturer).

End of life

The state of, or steps and process taken, when a material, consumable, or equipment is no longer useful to the manufacturer. Means or state of final disposal.

Energy source

The means or way by which energy is produced for consideration in facility inputs, flows, and balances.

Environmental assessment

The assessment of environmental consequences (positive and negative) of a plan, policy, program, or project.

Environmental impact (or burden)

The total effect of one or more of the materials, consumables, or equipment upon one or more of the identified environmental impact categories, usually during one phase of its life cycle, such as use.

Environmental impact assessment

Environmental assessment applied to actual projects. It is the process of identifying, evaluating, and mitigating the biophysical and social consequences of development proposals prior to commitments being made.

Equipment

The durable elements of a single-use or durable process train, e.g. a housing or support.

Functional unit

An operational constant or standard used to allow comparison of performance across many related processes. Often a productivity within a specified volume, dimension, or weight.

Hybrid process

A series of manufacturing operations composed of various combinations of disparate processing equipment or operations, such as of SU and durable.

Impact category

The environmental or natural systems damaged by an activity or emission. Often assigned to midpoint (physical or chemical change) or endpoint (where an adverse outcome occurs) such as “greenhouse gas production.” Synonyms include an environmental impact category.

LCA

A tool to assess the environmental impacts of a product, process or activity throughout its life cycle from the extraction of raw materials through to processing, transport, use, and disposal.

Life cycle impact assessment

An individual component of an LCA, assessing net effect of individual unit operations (or production subsystem) upon each impact category.

Life cycle impact

The cumulative environmental impact of a material, consumable, or equipment from cradle‐to‐grave or concept to disposal.

Life cycle stage

An individual unit operation (or production subsystem) and its product within the system boundary.

On‐site activities

Those steps performed in the facility of immediate discussion such as production of raw materials, biological products, or end‐of‐life activities. Synonyms include on‐site operations.

Process configuration

SU and durable components used in the manufacture of a product.

Process scale

Volume of the largest container or bioreactor in a manufacturing train or a metric of the material volume/mass of the process throughput.

Product type

Examples in biomanufacturing include cell‐based products (human mesenchymal stem cells, T‐cells, etc.) and biotherapeutics (therapeutic proteins, monoclonal antibodies, mAbs, hormones, growth factors, etc.), vaccines, chemicals, and waste processing.

SU process

A series of manufacturing operations composed of equipment having removable product contact elements made of non‐cleanable or reusable materials that are used once only. Synonyms include disposable technology.

13.3  The Evolution of SU Technologies

Table 13.1  (Continued) Term

Definition

Siting geographies

The physical location of the final product manufacturing facility.

Strategic environmental assessment

Applies to policies, plans, and programs most often proposed by organs of state.

SU materials

Non‐cleanable or reusable materials that are used only once.

Supplier

Manufacturer of bioprocessing equipment, consumables, or materials.

Supply chain

All the steps and activities employed in getting equipment, materials, or consumables from their manufacturer to end users.

Sustainability

Maintaining biological, social, and physical systems to be diverse and productive over time. Includes cultural preservation and social equity; economic and technical development; and resource and system conservation.

System boundaries

The entire scope of all processes involved in each operation, activity step, and product addressed in an LCA study. The sum of all addressed life‐cycle stages.

such reasons as a manufacturer’s concern for the environment, anticipation of growing environmental regulations, economic benefits from sustainable approaches, and social pressures from a concerned public. An application of basic green biological manufacturing principles can guide in the choice of biomanufacturing methodologies and processes to reduce the environmental impact. The essential principles of green manufacturing [5] provide an excellent starting point for establishing compliant and responsible biomanufacturing. But, efficiency in these approaches can only be accomplished through their adaption and focused implementation ­following an examination of the specific operations of ­biotherapeutics production methods. This is because an accurate assessment of the absolute and relative environmental impact from the employment of SU systems requires a rigorous and systematic science‐based approach. The several initial environmental analyses performed for SU biologics production illustrated the need for rational and comprehensive analysis techniques such as LCA to  understand the total, specific, and “cradle‐to‐grave” environmental strain imposed by both durable and SU approaches [6]. The first SU products employed in biotechnology included T‐flasks, pipettes, tubing, and filter membranes in the 1970s. Each device (or component part) back then was typically made from a single plastic resin, such as polystyrene or polypropylene, and, if we had been interested, would have had many end‐of‐life options including recycling. While SU versions of such small items as pipettes quickly overtook their durable precursors through the 1980s, it was not until the 1990s that larger containers and equipment housings became available in plastics. Disposable cartridge‐style filters and flexible‐walled large‐capacity, multiply ported bags for materials liquid storage and transportation rapidly overtook durables through the early 1990s. Sustainability

concerns were raised because of the order of magnitude greater mass of these larger components, as well as the fact that many became integrally composed of divergent types of plastic resin: e.g. filter housings were a different plastic than the filter membrane, and films (or webs) became multilayer laminates of different thermoplastic resins. At the turn of the century, early adopters began experimenting with such active consumables as SU bioreactors, mixers, and other large vessels and equipment we understand as the core of this technology today. Early on, many of the intermediate‐sized utensils existed as SU plastic inner components contained in a durable outer housing. However, of late we have even seen the development of larger single (and multiple)‐use outer plastic housings. Implementations rapidly expanded to support process harvest, collection, purification, and powder handling. So, the increased dimensions and scale of rigid SU housings of such equipment as filters, vessels, and chromatography columns allowed for the commercialization of even more SU equipment supporting more biomanufacturing operations. For the past five years, SU implementations of even the most exotic manufacturing appliances such as heat exchangers and large‐scale inoculum freezers have been introduced. So, we have seen a development in the types of devices available in SU, the specific applications or operations in which they are employed, and their materials of construction. Similarly, the specifications for these products as components of regulated processes have evolved. For example, early concerns for strength and closure integrity were eclipsed by interest in coextruded multi‐ply films displaying multiple properties such as puncture‐resis­ tance, low extractable profile, weldability, and gas barrier capability. Today, activities include standards and best practices being established for leachables, extractables, and particulate levels as well as evaluation of the utility of such new resins as specialty low‐density polyethylene,

171

172

13  Environmental Impacts of Single‐Use Systems

polyolefin, ethylene vinyl acetate mono‐material, fluorinated ethylene propylene, and fluoropolymer.

13.4 ­Implications in Sustainability The scale of SU applications and consequent mass of materials is the first consideration here. No one would be concerned about the use of SU forceps in brain surgery, as their incidence of use and total annual mass would be so small. While the mass of SU components in biomanufacturing is rather small in comparison to, e.g. foods packaging, the overall mass in all biotechnology is considerable and growing. The materials employed, by definition, tend to be rather biologically inert and nontoxic. Also, they are for the most part thermoplastics, which would (all things being equal) support creativity in end‐of‐life scenarios. However, all things are not equal: the multilayer l­ aminate nature of most of the films used in biomanufacturing, their contamination in use, and the diversity of plastics employed, has made recycling challenging. Current value recovery options seem to be limited to Current value to mixed‐plastic products like wood substitutes and incineration with energy recovery. After use in biomanufacturing, SU components are usually regarded as ­contaminated and possibly infectious. For example, a used SU bioreactor liner will be considered biologically contaminated and require sterilization by some means before further handling. Various local, state, and federal regulations may stipulate special steps required to prepare product‐contact materials for transport, storage, and recycling or disposal.

13.5 ­LCA – A Holistic Methodology We now turn to the question of how to decipher and interpret the environmental implications of SU technolo­ gies. One is often tempted to think of environmental impacts or benefits in terms of a single metric such as energy efficiency, water consumption, carbon emissions, or waste. While single‐metric perspectives are convenient and easy to grasp, in most cases they cannot capture the complexity or trade‐offs that are inherent in many technology selection or product design decisions. Another common error is to only consider such a metric during a discrete and arbitrary period of time. Over the past several decades, a methodology known as LCA has evolved in which environmental impacts or benefits of a product or technology can be evaluated across a broad spectrum of indicators. LCA is now an internationally recognized methodology [7, 8] that can be used to examine comprehensive environmental impacts across the full life cycle of a given product, from raw material extraction and refining through manufacturing, use, and end‐of‐life disposal or recycling (Figure 13.1). The focus on multiple life‐cycle stages allows one to gain insight regarding burden shifts from one life‐cycle stage to another, while the inclusion of multiple environmental indicators allows one to understand both the aggregate environmental impacts, as well as trade‐offs among different environmental categories. This results in a comprehensive view and a more accurate picture of associated environmental issues and improvement opportunities. Many types of distinct environmental concerns impacting the product life cycle are evaluated in LCA. The environmental indicators can be located at any

Minerales Respiratory Land use Heavy metals Ozone depletion Resources Bio diversity

More than just carbon footprint

Use

Human toxicity

Climate change Particulates SOx Eutrophication Acidification

Material processing

End of life

Waste Energy NOx Material scarcity Smog Water Ecotoxicity VOCs

Resources

Manufacturing

Distribution

Ecosystem quality

Natural resources

Figure 13.1  Schematic diagram of the product life cycle (left) and environmental impacts typically evaluated in LCA.

Human health

13.6  LCA Applied to SU Technologies

intermediate position (“midpoint,” representing a physical or chemical change in the environment) or can quantify the resulting damage (“endpoint,” where the adverse outcome occurs) in the cause‐and‐effect chain. A detailed LCA study can “peel the onion” of complex environmental effects, yielding a credible and transparent opportunity to compare different products and technologies that perform the same function, or to understand where environmental “hot spots” lie within a given ­product or technology.

13.6 ­LCA Applied to SU Technologies 13.6.1  Early Attempts to Examine the Environmental Aspects of SU Technologies Sinclair et al. [9] compared the environmental aspects of traditional bioprocessing (durable technology) with a facility implementing SU technologies for cell culture, mixing, holds, and liquid transfers for mAb production. While they did not apply a formal LCA methodology, they compared various metrics including water, material, and electricity usage, labor requirements, space, and carbon footprint, suggesting that a shift to SU technologies could potentially result in lower material consumption and carbon emissions. Rawlings and Pora [10] performed comparative energy consumption calculations that included sterilization, cleaning, and materials, suggesting that the energy consumption of SU bioprocessing might be lower than for traditional bioprocessing. These studies, while not comprehensive, provided the first indications that the environmental implications of SU technologies might not be as bad as initially feared.

­ rocess trains comprising 14 unit operations from N‐2 p seed fermentation through product purification and considered the entire life cycle for both types of bioprocess technologies, including supply chain, use, and end of life. Supply chain included materials and manufacturing of all process equipment and consumables needed to support a 10‐batch campaign, including presterilization of SU components. Use stage included all activities that occur during production, including sterilization and cleaning of equipment between batches for  the traditional process train. End of life included ­transport of consumables and equipment to end‐of‐life treatment, disposal of consumables, and disposal, reuse, or recycling of durable equipment. The results of this study indicated that the SU process train exhibited lower environmental impacts compared to the traditional approach in each of the 18 environmental impact categories studied. The finding was attributable primarily to the reduced requirements of SU technologies for steam, energy, and water‐for‐injection, particularly during the cleaning and sterilization steps that are required between batches when using traditional process technologies. The study provided a variety of further insights, such as: ●●

●●

●●

●●

13.6.2  LCA Applied to SU Technologies The first application of LCA methodology to SU process technologies was a screening (high‐level, limited scope) study comparing SU and traditional bioreactors for mAb production at a 500 l working volume [11]. Although the bioreactor is just one unit process amidst the full process train, the results of this study suggested that a life‐cycle perspective can reveal a more complex series of environmental trade‐offs that are not reflected by a focus on waste streams. In fact, the study showed that operational aspects and manufacturing of the SU components have a far greater effect on environmental impacts than the process waste streams. The first comprehensive, detailed cradle‐to‐grave LCA study of traditional vs. SU technologies, performed by GE Healthcare, compared the production of mAb over a 10‐batch campaign at three production scales (100, 500, and 2000 l) [12–15]. The study evaluated the entire

A substantial majority of environmental impacts occurred during the use (operation) stage for both traditional and SU technologies. Carbon and energy impacts in the supply chain were slightly higher for SU compared to traditional due to the increased manufacturing activity required to supply the SU consumable components, but supply chain impacts were relatively low compared to the use stage. Water usage was lower for SU compared to traditional technologies in all life‐cycle stages: supply chain, use, and end of life. Environmental impacts from end‐of‐life disposal were higher for SU compared to traditional, but end‐of‐life impacts were almost negligible in the context of the overall life cycle of either SU or traditional processes.

The study was particularly effective at revealing complex layers of information that would not have been revealed through a focus on single environmental metrics. It revealed burden shifts (e.g. shift of environmental impact from use stage into supply chain) and trade‐offs (e.g. slight increases in end‐of‐life disposal impacts significantly offset by energy and water savings during use stage). More importantly, this study unequivocally showed that SU process technologies can be less environmentally impactful compared to traditional approaches. While end‐of‐life waste management considerations are higher for SU, when examined in a setting of the overall life cycle, the burden of waste management is negligible relative to the aggregate environmental benefits of SU process ­technologies. The results seemed counterintuitive to our

173

174

13  Environmental Impacts of Single‐Use Systems

Table 13.2  Parameters for two GE LCA studies. 2010–2012 Study

2016–2017 Study

Molecule or particle to be manufactured

mAb only

mAb Adenovirus vaccine (AdV)

Process technology

100% SU retrofit 100% stainless steel (SS) traditional

100% SU retrofit 100% SS traditional Hybrid (62% single‐use)

GE Healthcare SU products

WAVE Bioreactor ReadyToProcess fluid management portfolio

WAVE Bioreactor Ready ToProcess fluid management portfolio Xcellerex XDR bioreactor Xcellerex XDUO mixers HyClone portfolio ÄKTAready chromatography system

Geography

US Average

Boston, Massachusetts, USA California, USA Sao Paulo, Brazil Istanbul, Turkey Shanghai, China Dortmund, Germany Cork, Ireland

End‐of‐life operations

Interaction

Autoclave‐landfill Shred‐autoclave‐landfill Incineration with energy recovery Recycling

Campaign type (mAb)

6 g/l 10‐batch campaign

6 g/l 10‐batch campaign

Process scales

mAb 100 l mAb 500 l mAb 1000 l

mAb 200 l mAb 500 l mAb 1000 l mAb 2× 2000 l mAb 4× 2000 l mAb 2000 l AdV 200 l AdV 500 l

­ revious uninformed speculation and led to a changed p environmental perspective on SU processing technology within the industry. Meanwhile the biopharmaceutical industry has continued to evolve in terms of global expansion of biomanufacturing capacity, increased pressure to maximize utilization while minimizing area footprints, and the introduction of newer SU process technologies. These drivers, combined with increasing interest in sustainability within the biopharmaceutical manufacturing sector, led GE Healthcare to launch another detailed LCA study that focused on a more thorough understanding the environmental impacts of SU process technologies [6]. This second study was performed with data gathering assistance from biopharmaceutical manufacturers. The study looked at both mAb and vaccine production, incorporated some of the latest SU product technologies, considered regional impacts for emerging markets, and explored a more comprehensive array of end‐of‐life treatment options for SU components. It focused on

impacts on climate change, energy, and water, as well as several other aggregated categories in which a comprehensive array of impacts were grouped with respect to their damage to ecosystem quality, human health, or natural resources [16]. The scope of this study compared to the previous study is shown in Table 13.2. This second study looked at many scenario permutations  –  approximately 630 combinations of product (mAb vs. adenovirus vaccine), process technology (traditional, SU, and hybrid), geography (7 locations), and 5 end‐of‐life treatment options. Representative results are used to convey key insights. Figure  13.2 compares the impacts in five categories for traditional and SU processing of mAb at a 2000 l scale for a bioprocessing facility located in Boston. The results show that the SU process configuration exhibited lower life‐cycle environmental impacts in all categories, once again primarily due to significant reductions in CIP/SIP support and less need for cleaning/sterilization between batches. A burden shift occurs in that supply chain impacts are higher for SU

100 90 80 70 60 50 40 30 20

Climate change

Human health Supply chain

(manufacturing and distribution of SU components). The exception is water consumption in which there is no burden shift of water impacts from use stage to supply chain when SU process technology is adopted. End‐of‐ life impacts are negligible compared to supply chain and use‐stage impacts. The effect of biopharma manufacturing facility geography is interesting and nuanced. Figure 13.3 shows that the two most impactful geographic variables related to climate change are: (i) how “green” the electricity grid is in each geography and (ii) transport logistics – the environmental impact of shipping the SU components by air, land, or sea, which also depends on the biomanufacturing facility geography relative to the point of SU component manufacturing and distribution [17]. As shown in Figure  13.3, the difference between SU and traditional decreases as electricity grids become cleaner; while energy consumption is still higher for traditional vs. SU, the environmental impact of energy consumption is less pronounced. Transport logistics are also quite important. For example, the comparative results for facilities in Sao Paolo are dominated by transport logistics with no clear benefit to either process technology since energy consumption has less environmental impact (more than 70% of electricity in this region comes from hydropower). Note that when considering freshwater consumption, SU is always better than traditional regardless of geography, electricity grid, or transport logistics [6]. Figure 13.4 shows a comparison of life‐cycle impacts in supply chain, use, and end‐of‐life stages for adenovirus vaccine (AdV) production by traditional or SU for a facility

Ecosystem quality Use

Resource consumption

Single-Use

Traditional

Single-Use

Traditional

Single-Use

Traditional

Single-Use

Traditional

0

Single-Use

10 Traditional

Figure 13.2  Impact comparison for traditional, hybrid, and single‐use technologies at 2000 l scale in Boston (assuming autoclave/landfill in end‐of‐life).

Percent of highest impacting configuration (%)

13.7  Sustainability Efforts in the BioPharma Industry

Water consumption

End-of-life

located in Boston, USA. The SU process technology shows a lower environmental impact in most cases, with a pattern of results quite similar to what was found for mAb. These two LCA studies together suggest several key insights: ●●

●●

●●

●●

SU technologies often result in reduced life‐cycle ­environmental impacts. Disposal of SU components at end of life does not contribute significantly to net life‐cycle environmental impact. Traditional processes are affected most by CIP/SIP and water for injection (WFI) energy use, whereas SU processes are most affected by distance/mode of transport. Region of choice for facility does contribute significantly to environmental impact due to transport logistics and power grid differences.

13.7 ­Sustainability Efforts in the BioPharma Industry In 2005, the American Chemical Society (ACS) formed the Green Chemistry Institute Pharmaceutical Roundtable (GCIPR) to improve the sustainability of biomanu­ facturing through incorporation of green chemistry and related principles into biologics development and manufacturing. GCIPR is working on methodologies and insights to improve the industry’s environmental footprint in process development, cleaning science, and facilities operations [18]. Another ACS group, the

175

13  Environmental Impacts of Single‐Use Systems 1000000

800000 Climate change (kg CO2-eq)

600000

400000

200000

Lo gis tics Hig dis tan h ce Lo w

0

Traditional Single-Use Traditional Single-Use Traditional Single-Use Traditional Single-Use Traditional Single-Use Traditional Single-Use Traditional Single-Use

Traditional

High

Single-Use

Dortmund

Low

Electricity grid impact

Cork

Istanbul

Boston

Los Angeles

Shanghai

Sao paulo

Figure 13.3  Comparison of climate change impact (y‐axis) between traditional and single‐use mAb bioprocess by siting geography, ranked by logistics distance (z‐axis) and electricity grid impact (x‐axis). On the x‐axis, the geographies are arranged left‐to‐right in terms of the “green‐ness” of the electricity supply in each region. On the z‐axis, the geographies are arranged front‐to‐back in terms of transport logistics intensity. Data are for a 2000 l mAb process. End‐of‐life disposal is autoclave followed by landfill.

100

80

60

40

Climate change

Human health

Supply chain

Ecosystem quality

Use

Resource consumption

Single-Use

Traditional

Single-Use

Traditional

Single-Use

Traditional

Single-Use

Traditional

0

Single-Use

20

Traditional

Percent of highest impacting configuration (%)

176

Water consumption

End-of-life

Figure 13.4  Potential environmental impact of AdV production for traditional and SU process configurations with 50 l production volume, sited in Boston.

13.8  End‐of‐Life (Waste) Management

Committee on Environmental Improvement (CEI), is working toward a vision of “a sustainable world enabled through the sustainable practice and use of chemistry…” by advancing “…sustainability thinking and practice across ACS and society for the benefit of Earth and its people.” The Bio‐Process Systems Alliance (BPSA) formed its  Sustainability Committee (http://bpsalliance.org/ committees) to provide information about the environmental impact of single‐use technology (SUT) and focus member company enthusiasm on this topic. The committee’s rapid growth demonstrates that sustainability is a priority for BPSA. Team accomplishments so far include summaries of available studies, a list of existing recycling/reuse programs, visits to recycling facilities, publication of relevant studies and resources, and an investigation of related efforts by similar groups. Notably, many SU materials suppliers and biomanufacturers are taking significant steps individually and collectively to study the issue, support the design of solutions, and reduce the environmental impact of their operations.

13.8 ­End‐of‐Life (Waste) Management “End‐of‐life” considerations are a big part of most discussions about SU and the environment because of the visibility of the pre and post‐process material. As discussed above, LCA studies show that post‐process handling of SU materials has relatively little influence on environmental impact. However, management of these materials is operationally important and is often a primary concern when casually examining the process. SU materials require handling by manufacturing personnel, storage costs, and disposal logistics that are not a part of traditional operations. Post‐process SU materials may be a substantial fraction of a manufacturing facility’s solid waste output. The conventional waste minimization paradigm “Reduce–Reuse–Recycle” can be put to good use in evaluating the material inputs and outputs in the supply, use, and disposal of SU materials. SU packaging, and the SU components themselves, present “reduce” and “reuse” opportunities for designers. “Recycle” is also an opportunity, but execution of SU material recycling programs are affected by how the SU materials are used, the location of the biomanufacturing facility, and the relationship of SU materials to other waste streams. SU components are manufactured in facilities that are rarely colocated with the biomanufacturing consumers. SU components may need to be shipped hundreds or even thousands of kilometers. Transportation impacts on sustainability have been described above. In the

“end‐of‐life” discussion, SU packaging required for safe transportation is a distinct and significant consideration. By reducing packaging (overbags, cartons, protective foam, etc.), SU suppliers will use less material and less fuel, and they will reduce storage space requirements throughout the supply chain. Most visibly for the consumer, packaging material handling, storage, and discard will be reduced. Customers receive SU components as palletized cartons. These pallets need to be received, stored as intact palletized units, or unpacked and stored as individual cartons. Cartons may contain one or several SU components. Each assembly is sealed in polyethylene bags, and it is usual to have a second bag sealed around the first. Prior to biomanufacturing use, protective packaging and the bagged assemblies are removed from the cartons. The protective packaging is almost always composed of SU materials. Packaging is segregated for recycling or mixed into a common facility waste stream. The SU components are transferred to airlocks leading to manufacturing areas. The polyethylene outer bag is removed, and the single‐bagged assembly is moved into the next cleanroom. At the point of use, the assembly is removed from its bag. The bag is discarded, and various packaging aids such as foam tape, bubble wrap, and cable ties used to secure items in transit are removed and discarded. Complicated assemblies such as bioreactors and SU mixers include additional packaging to prevent damage to sensitive parts during transit, storage, and setup. At this point in the process, there has been no biomanufacturing, but there has been a substantial quantity of material discarded. Pallets, cardboard cartons, polyethylene bags, and packing foam are all available for materials recovery and none of these materials is likely to be, or to be considered as, contaminated. There are many options for “reduce–reuse–recycle” of SU system packaging. After the SU actual components are used in biomanufacturing, the operation stops and the SU components are uninstalled. Consideration needs to be given to any product residues that are in or on the SU components. Is the material considered hazardous per company rules or local regulations? Is a decontamination needed prior to removal from the facility? If not, is decontamination needed prior to any reuse or recycle? Does the necessary decontamination method interfere with value recovery? After any required decontamination, the SU components are moved out of the production facility to a storage area. This area is managed according to company and government waste regulations including adherence to holding time limits and record keeping requirements. From a materials management perspective, “reduce” is the primary target for improving end‐of‐life strategies for SU materials. Materials reductions are supported by improved packaging designs, avoidance of overbuying,

177

178

13  Environmental Impacts of Single‐Use Systems

and better use of standard SU component designs. Standardization promotes improved inventory management as the same SU component can be used in multiple applications. Suppliers support their customers with fewer stock‐keeping units (SKUs), order fulfillment times are reduced so that more sustainable shipping options can be used, and customers manage fewer SKUs and discard less expired inventory. SU material post‐process handling will be dependent on the nature of the production operation and the products used, on geography, and on user preferences for waste management. Those users who are in close proximity to waste‐to‐energy facilities (see, for example, http://bpsalliance.org/wp‐content/uploads/2017/07/10‐ VANBRUNT.pdf ) or to recyclers who manage infectious waste (see, for example, https://www.triumvirate.com/ red2green‐production) should give careful consideration to these options over landfilling or incineration without energy recovery. Those users who would have to ship post‐process SU materials over long distances may have to look at the big picture and do the least harm by using a less beneficial end‐of‐life option.

13.9 ­Summary and Conclusions A primary benefit of science‐based approaches such as LCA is a comprehensive understanding of environmental impacts of a scenario throughout the entire life cycle. The detailed LCA studies performed thus far have revealed layered insights that provide useful guidance to biomanufacturers as they assess their strategic technology development and expansion plans in this age of an increasingly regulated and resource‐constrained operating environment. Two key insights that have emerged from these studies are that SU technologies can be less environmentally impactful than traditional biomanufacturing approaches, and that environmental impacts related to disposal of SU components at end of life are negligible in the context of the cradle‐to‐grave life cycle of biomanufacturing. By cataloging every aspect of the process, and objectively comparing the environmental detriment or benefit in all relevant impact categories, we can more effectively focus industry efforts to improve what environmental pressures remain, and drive toward a more sustainable future.

Nomenclature ACS AdV BPSA CEI CIP GCIPR ISO

American Chemical Society Adenovirus vaccine Bio‐Process Systems Alliance Committee on Environmental Improvement Cleaning in place Green Chemistry Institute Pharmaceutical Roundtable International Organization for Standardization

LCA Life cycle assessment mAbs Monoclonal antibodies SIP Sterilization in place SKUs Stock‐keeping units SS Stainless steel SU Single‐use WFI Water for injection

­References 1 N.U. (2018). Unintended consequences. http://www.

2

3

4

5

pollutionissues.com/Te‐Un/Unintended‐Consequences. html#ixzz5D4zalkND (accessed 3 August 2018). N.U. (2018). The sustainability problem. http://www. developforthelongterm.com/the‐sustainability‐problem. html (accessed 3 August 2018). Junker, B. (2010). Minimizing the environmental footprint of bioprocesses, part 1: introduction and evaluation of solid‐waste disposal. BioProcess Int. 8 (8): 62–71. Junker, B. (2010). Minimizing the environmental footprint of bioprocesses, part 2: evaluation of waste water, electricity and air emissions. BioProcess Int. 8 (9): 36–46. Anastas, P. and Zimmerman, J. (2003). Design through the twelve principles of green engineering. Environ. Sci. Technol. 37 (5): 94A–101A.

6 Flanagan, W. (2016). Single‐use and sustainability:

quantifying the environmental impact. https://www. bioprocessonline.com/doc/single‐use‐technology‐and‐ sustainability‐quantifying‐the‐environmental‐impact‐ 0001?vm_tId=1969681&user=54aa44f0‐9ab6‐444d‐ 9995‐53255bb6bd74&utm (accessed 3 August 2018). International Organization for Standardization (2006). 7 ISO 14040 (2006): Environmental Management – Life Cycle Assessment – Principles and Framework. Geneva: ISO. International Organization for Standardization (2006). 8 ISO 14044 (2006): Environmental Management – Life Cycle Assessment – Requirements and Guidelines. Geneva: ISO. Sinclair, A., Leveen, L., Monge, M. et al. (2008). The 9 environmental impact of disposable technologies. BioPharm. Int. 21 (11): s4–s15.

­  References

10 Rawlings, B. and Pora, H. (2009). Environmental impact

11

12

13

14

of single‐use and reusable bioprocess systems. BioProcess Int. 7 (2): 18–25. Mauter, M. (2009). Environmental life‐cycle assessment of disposable bioreactors. BioProcess Int. 8 (4): 18–28. Pietrzykowski, M., Flanagan, W., Pizzi, V. et al. (2011). An environmental life cycle assessment comparing single‐use and conventional process technology. BioPharm. Int. 24 (S11): 30–38. Flanagan, W. (2015). An environmental life‐cycle assessment: comparing single‐use and traditional process technologies for Mab production. BioProcess Int. 13 (11i): 10–26. Flanagan, W., Pietrzykowski, M., Pizzi, V. et al. (2014). An environmental life cycle assessment of single‐use and conventional process technology: comprehensive environmental impacts. BioPharm. Int. 27 (3): 40–46.

15 Pietrzykowski, M., Flanagan, W., Pizzi, V. et al. (2013). An

environmental life cycle assessment comparison of single‐ use and conventional process technology for the production of monoclonal antibodies. J. Clean. Prod. 41: 150–162. 16 World Health Organization. (2005). Ecosystems and human well‐being: a report of the Millennium Ecosystem Assessment. http://apps.who.int/iris/bitstream/ handle/10665/43354/9241563095.pdf;jsessionid= 9F90A4BDE3F3F1B0BDA88C7EC6330820?sequence=1 (accessed 3 August 2018). 7 IPCC (2007). Climate Change 2007: The Physical 1 Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press. 8 Budzinski, K., Ho, S., Millard, J. et al. (2015). Toward 1 sustainable engineering practices in biologics manufacturing. BioProcess Int. 13 (11i): 1–9.

179

181

14 Design Considerations Towards an Intensified Single‐Use Facility Gerben Zijlstra, Kai Touw, Michael Koch, and Miriam Monge Sartorius Stedim Biotech GmbH, Göttingen, Germany

14.1 ­Introduction The global biologics market is dynamic and the industry must alter the nature of its biomanufacturing networks in order to reflect the changing needs of its customers. Whereas much of the industry’s growth has been based on the tremendous success of monoclonal antibody (mAb) therapies, biological product pipelines being developed in response to unmet clinical needs currently are far more diverse than in the past and include non‐ antibody recombinant proteins, antibody–drug conjugates, and cell and gene therapies. Biopharmaceutical companies are increasingly recognizing that there are opportunities for growth in emerging markets in which they have not previously been present. This requires them to improve their global reach and potentially manufacture products within the region they are targeting. While this expansion is good news for the sector, firms must address the ever‐increasing pressure to decrease the costs of their medicines as payers reject high‐priced drugs and look toward generic versions of biological products to reduce market prices. These market dynamics are increasing the uncertainty faced by biotechnology companies. It is becoming harder for them to gauge the relative success of new products or the variation in demand they might experience in these new markets. Increased competition is compounding these difficulties as the market matures and as new entrants see opportunities to take advantage of the market flux. The biopharmaceutical industry is challenged to improve its manufacturing process robustness and reduce the costs associated with limited process understanding and control, e.g. causing suboptimal processes or even batch failures. It must decrease the time it takes to bring new production facilities up and running and to increase its speed to market. Facilities within future biomanufacturing networks must enable a significant reduction in the time it takes to switch over production

between different products and be able to respond to fluctuations in demand. No “one‐size‐fits‐all” facility‐type will meet all these future needs. Rather, biopharmaceutical companies will incorporate a range of different facility‐types into their production networks. The BioPhorum Operations Group facilities roadmap outlines six biomanufacturing scenarios ranging from large‐scale stainless‐steel fed‐batch operations, intermediate‐scale single‐use perfusion plants, intermediate‐scale multiproduct, single‐use fed‐batch facilities, small‐scale portable facilities, and small‐scale factories for personalized medicines. Scenarios in which smaller product volumes are required are more likely to be single‐use and distributed, while at the other end of this spectrum the larger facilities will be stainless steel and centralized [1, 2]. This chapter will focus on intermediate‐scale single‐ use facilities with continuous or intensified bioprocesses that can bridge the gap between large‐scale stainless‐ steel facilities containing 10 000 l bioreactors or larger, and intermediate‐scale single‐use facilities operated in regular fed‐batch mode. This is an area in which we are currently observing a significant amount of innovation and debate within the industry. It is the authors’ intention to show what these particular intensified single‐use biomanufacturing facilities could look like.

14.2 ­Moving Towards Intensified and Continuous Processing Producing biopharmaceuticals in batches rather than continuously has been common in the industry over the past couple of decades. This, despite the fact that perfusion cell culture was used at the industries’ inception as a method of producing labile biologics that would degrade if left unpurified in cell culture supernatant for prolonged periods of time. The productivity of fed‐batch cell ­cultures is limited because the cells do not reach their

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

182

14  Design Considerations Towards an Intensified Single‐Use Facility

maximum viable cell density until late in the culture duration. During the early stages of the culture, there are relatively few cells and these are allocating much of their metabolic energy to dividing rather than on synthesizing product. Engineers can increase the productivity of fed‐batch cultures if they seed the production bioreactor at very high cell densities using a seed bioreactor operated in perfusion mode. In these high inoculation fed‐batch ­circumstances, the viable cell density of the cultures can be limited by the accumulation of toxic metabolic by‐ products. It is possible, however, to remove these toxic components by also operating the main bioreactors using perfusion principles. In concentrated fed‐batch mode, a filtration device is used that allows the removal of spent cell culture media while retaining the cells and product. In perfused fed‐batches, the approach is the same, but the product passes the retention system (typically a microfilter) into the permeate flow. Bioreactors operated in ­concentrated fed‐batch and perfused fed‐ batch mode can generate extremely highly viable cell densities and ­ consequently, very high (cumulative) ­product titers. Concentrated fed‐batch cultures can be processed similar to traditional batch mode but with much higher output, such as 20–30 kg mAbs from a 2000 l bioreactor. By running a number of these bioreactors at the 2000 l‐scale in parallel and scheduling them in a way that operators harvest them every few days, the output of a single‐use ­facility can be greatly increased to allow commercial quantities of biologics to be supplied to the market. Perfused fed‐batch cultures can produce the same output, but typically require at least a partially continuously ­operating downstream processing (DSP) to capture and ­further process the product from the permeate flow. Bioprocess engineers will achieve the highest volumetric productivities in single‐use bioreactors by operating them in continuous perfusion mode for continuous biomanufacturing. In this way, they will generate and maintain the highest average viable cell densities over time as growth phases with low cell densities are eliminated. Operating cell cultures reliably using perfusion principles in any of the mentioned intensified fed‐batch and continuous perfusion cell culture modes requires relatively sophisticated equipment setups and control s­ ystems. To reliably maintain the long‐term (typically 30–90 days) continuous perfusion cultures, both from a logistical as well as from a technical perspective, is typically considered an additional risk especially by companies with a ­ ioprocess batch‐­processing history. Furthermore, many b experts think currently that continuous ­perfusion cultures take longer to validate than (intensified) fed‐batch processes and this could reduce the speed with which developers can bring new products to market.

The industry is applying continuous production t­echniques to the DSP steps used to capture and purify proteins because of their high volumetric productivities (see also Chapter  9). Indeed, as the productivity of upstream processing (USP) steps increases, the need for approaches that debottleneck purification trains will become even more urgent. Technologies such as multicolumn chromatography are delivering over fivefold improvement of specific productivity during protein purification [2]. Replacing purification resins with membrane adsorbers improves this productivity still further [3]. In practice, running perfusion cell cultures that are purified by batch‐purification processes is a widely adopted approach for the commercial manufacturing of complex and labile proteins. The authors are aware of a number of biopharmaceutical companies that are evaluating or at the early stages of implementing hybrid‐­continuous process solutions, where they intend to operate part of the production train in a continuous mode. This includes the purification of fed‐batch or ­perfusion ­cultures using multicolumn chromatography. A number of leading biopharmaceutical companies are exploring fully integrated continuous processing throughout the bioprocess but this is most likely a longer‐term objective. It is the opinion of the authors that while this approach will eventually provide the best product quality and the ­lowest production footprint, for the time being, the intensification of single‐use facilities will give the best compromise allowing rapid implementation, low cost of goods (COGs), and greatest flexibility. We envisage that biomanufacturing facilities based on six 2000 l single‐use bioreactors producing 10 g/l of protein will allow over 1000 kg/yr of product to be manufactured at a fraction of the cost it takes to build and operate a traditional “six‐pack” stainless‐steel facility with 15 000 l bioreactors operated in fed‐batch mode. A good example of a current state‐of‐the‐art facility recently built in Asia is a hybrid continuous plant which cost the company around $160 million [4]. It utilizes single‐use bioreactors with a maximum size of 2000 l each. Operated in fed‐batch mode, these bioreactors can produce approximately 5 g/l of antibody but concentrated fed‐batch processes could provide well over 10 g/l. The facility uses ballroom principles and is designed modularly in such a way that it could be replicated anywhere else in the world according to market needs. It produces around 1 metric ton of product per year but is 75% smaller than comparable traditional stainless‐steel facilities and has a footprint of approximately 120 000 square feet. The plant took around 15 months to be c­onstructed and required only a quarter of the capital investment compared to what a traditional facility would need. The company estimates that the annual operating expenses will be threefold lower than conventional designs that produce the same quantity of product.

14.3  Methodologies for Continuous and Intensified Single‐Use Bioprocessing

14.3 ­Methodologies for Continuous and Intensified Single‐Use Bioprocessing Companies that provide single‐use technologies to the biopharmaceutical industry will have portfolios of ­products and services that support the intensification of bioprocesses. Figure 14.1 shows how suppliers are able to integrate single‐use technologies to create end‐to‐end process platforms for the production of mAbs, recombinant vaccines, and gene therapy products. Critically, engineers that are purchasing process‐scale equipment should consider the scalability of systems and whether tools exist that will allow early‐stage p ­ rocess development

to be performed effectively so that, it can later be translated to large scale processes. Vendors have specifically engineered high‐throughput automated process development tools that allow the development of the sophisticated control strategies (using a “Design of Experiments” approach) required for handling highly productive upstream and downstream unit operations. Process engineers should select single‐use N‐2, N‐1, and production bioreactors based on their ability to provide the ­ oxygen mass transfer needs of high cell‐density ­cultures and likewise, they will require clarification technologies that can handle the high solid content of high inoculation and concentrated fed‐batch cultures. The operation of these challenging processes will need supporting by the appropriate automation and control

Intensified upstream process Cell bank

Rocking motion or stirred tank bioreactors in perfusion

Single-use systems for media and buffer preparation and storage

Single use stirred tank with PAT tools for (concentrated) fed-batch

Single-use centrifugation

Body feed filtration

Liquid Media

Sterile and nano-filtration for media and buffers

Seed expansion

Media preparation

Production

Cell removal

Medium, feeds

Intensified Downstream Process

Process control and monitoring systems

Monitoring

Membrane adsorbers

Low pH

Affinity chromatography Clarified harvest

Buffer preparation

Capture

Virus inactivation

Virus filtration

Cation exchange chromatography Bind and elute

Buffer

Single-use systems for buffer preparation and storage

Figure 14.1  An intensified single‐use process platform.

Sterilizing (Pre)filters

Buffer exchange and formulation Flow-through

Virus removal

UF ⃒DF

Sterile filtration

Filling

183

184

14  Design Considerations Towards an Intensified Single‐Use Facility

systems that provide the highest levels of process control, operator guidance, and integration capabilities. The development of online process analytical technologies (PAT) in combination with multivariate data analysis give rise to advanced process‐control strategies. The ultimate goal is to combine the data generated by these refined technologies and mechanistic modeling toward predictive control loops. In DSP, firms should invest in technologies that are capable of processing large quantities of protein products at high flow rates and able to withstand the dynamic process conditions experienced during intensified processing regimes [5].

14.4 ­Process Development for Intensified Biomanufacturing Facilities The starting point for the design of any facility is the manufacturing process itself. The selection of the appropriate processing technologies will depend on the ­characteristics of the biologic and the annual production requirements needed to satisfy market demand. Engineers can use process‐modeling software, such as BioSolve Process to run simulations of different process configurations. They can use these in turn to formulate equipment lists that allow capital expenditure (CAPEX) requirements to be estimated and the likely COGs from the facility to be determined. Armed with this information, process development scientists can direct their efforts toward specific development objectives that will ensure their intensified process will use the available facility most efficiently. Table 14.1 shows four different process configurations that were evaluated in BioSolve Process to explore the best way of intensifying a mAb production process. Scenario 1 is a baseline process with a 2000 l, classical fed‐batch cell culture expressing product to 3 g/l and three standard resin chromatography steps. Scenarios 2–4 simulate processes with different operation modes in USP using different perfusion strategies that achieve final product titers up to 10 g/l. For

clarification, body feed filtration was used to process the harvest of scenarios 2–4 and for purification a classical three‐step chromatography approach was used. In all cases, the 6× 2000 l bioreactors are operated in a staggered mode and the harvest is purified by a single DSP train for maximal asset utilization and minimal COGs. Per bioreactor maximally 20 runs per year have been defined, to allow for facility maintenance and (limited) product changeover. The results of the modeling work are shown in Figure  14.2 and show that switching to a concentrated fed‐batch culture more than triples the yearly throughput (i) of the facility but also increases the COGs per gram by approximately 10% (ii). The main driver of the COGs in the concentrated fed‐batch scenario is the medium (iii). Reducing the medium consumption has a positive effect on the COGs when using concentrated fed‐batch. When looking at the N‐1 perfusion scenario, one can see that the COGs per gram are the lowest of all four scenarios, while doubling the overall annual throughput in comparison to the classical fed‐batch scenario. The output of the models provides actionable information that allows process development scientists to develop intensified cell cultures that have greater productivities without the corresponding increase in cost. Each simulation generates not only the cost associated with each step but also a generic mass balance report. Using this COGs modeling approach to analyze the different upstream operation modes also yields valuable information for the design and engineering of a flexible manufacturing facility. As all these scenarios make use of the same facility setup, the manufacturing strategy can be decided based on process development data, while there is no need to change the layout of the facility.

14.5 ­The Intensified Biomanufacturing Facility Following the process modeling scenario, a process specific or even a generic process mass balance for ­ ­multiproduct has to be defined as a fundamental basis

Table 14.1  Four monoclonal antibody process scenarios modeled in the BioSolve Process. Estimated upstream titer (g/l)

Scenario

Operation mode

Feed rate

1

Fed‐batch

NA

3

120

2

N‐1 perfusion

1 VV/d

6

120

3

Concentrated fed‐batch

1 VV/d

10

120

4

Concentrated fed‐batch

0.5 VV/d

10

120

NA-not available; VV-vessel volume.

No. of batches

14.5  The Intensified Biomanufacturing Facility

(a)

(b)

Throughput kg/yr

60 50 40 30 20 10 0

2000 1500 1822

1000 500

1822

1094

547

0 Fed-batch

N-1 perfusion

Conc. fed batch

Conc. fed batch (50% media reduction)

(c)

Cost per gram EUR/g

52

42

Fed-batch

N-1 perfusion

58

47

Conc. fed batch (50% media reduction)

Conc. fed batch

COG’s overall distribution

120 000 000 100 000 000 80 000 000 60 000 000 40 000 000 20 000 000 0 Annual cost (EUR) Fed-batch

Capital charge

N-1 perfusion

Consumables

Materials

Conc. fed batch

Labor

Others

Conc. fed batch (50% media reduction)

Figure 14.2  Model outputs for the different mAb‐process scenarios as described in Table 14.1.

for the technology selection and sizing of the equipment. The mass balance contains the relevant processing parameters for all process steps and defines the required volumes and processing times (Figure 14.3). After this exercise, the individual process steps can be defined within certain unit operations and the process flow can be represented either through a block flow diagram or directly through a process flow diagram. This is typically performed during conceptual and basic design phase respectively. During the facility design phase, the main focus is on the process and its requirements. The processing suites will be designed around the process and not vice versa (Figure 14.4). In order to help biopharmaceutical companies address key market challenges, a generic process platform, like Sartorius’ P4S “Process for Success”, based on single‐use technology for mAb processes, is a good starting point for the ensuing conceptual design and basic design activities for the intensified manufacturing facility. The design of the facility is highly flexible enabling multiple products based on a variety of manufacturing scenarios to be produced from the same plant with short process changeover times. Shorter changeover times at lower

cost are enabled by the fully single‐use nature of the USP process equipment, the use of easy to exchange single‐ use consumables for hard to clean system components in hybrid single‐use DSP and finally, the functionally closed operations. This prevents contamination of the cleanroom surroundings and further DSP and even allows for ballroom operation for the integrated process through to virus filtration. Ballroom operation further enhances the process flexibility, by enabling easy exchange of unit operation equipment or change of unit operation sequence. This flexibility means that the capacity can be modulated readily in order to produce according to forecasted market demand of a new drug and based on early‐stage development data. If we take the example of a 6× 2000 l single‐use plant, capable of using different upstream manufacturing ­scenarios, we estimate the maximum capacity will be between 500 and 2000 kg/year, depending on the mode of operation (Table  14.1 and Figure  14.2). The related capital investment cost needed for the facility will be in the region of EUR 100 million and the annual costs will be a fraction of those of a traditional stainless steel ­facility, which are dominated by depreciation.

185

186

14  Design Considerations Towards an Intensified Single‐Use Facility

Mass balance report IFOF2000 Input

In (l): Growth media

In (l): Material in (l): Media start Feed media Base Antifoam

In (l): Material in (l): Media start Feed media Base Antifoam

In (l): Material in (l): Media start Glucose Feed media Base Antifoam

Unit operation

5 20

25 74.5 70 0 2.5 2

100 397.5 380 0 7.5 10

500 1391 1 10 1380 0 0

M01

Output

Cell seed cultivation (N-3) 01 Shake flask Batch duration

h

Volume (l):

25

3.0

Cell seed cultivation (N-2) 02 CultiBag RM M01 M02 B01 B02

Bioreaction volume

l

Batch duration

h

Volume (l):

100

200.0 3.0

Cell seed cultivation (N-2) 03 CultiBag STR M01 M02 B01 B02

Bioreaction volume

l

Batch duration

h

Volume (l):

500

500.0 3.0

Cell cultivation 04 BIOSTAT STR M01 M06 M02 B01 B02

Batch duration

day

10.0

Expected product titer (g/l)

g/l

10.0

Sizing factor

%

0%

Volume (l):

2000

Concentration (g/l):

10

Figure 14.3  Excerpt of mass balance sheet.

The detailed design and realization of this kind of an intensified facility may require some adaption of a ­number of processing units and technologies as there may be certain limitations with respect to the scale of the currently available single‐use products and technologies. However, these limitations are minor and will no doubt be overcome in the near future, which will enable

full‐scale intensified processing within single‐use ­biomanufacturing facilities. The operation of this type of single‐use intensified bioprocess needs careful consideration of both facility room layouts and a suitable automation and process control system. Process and facility requirements are key business drivers and their consideration forms an important

14.7  Intensified Upstream Processing

part of overall project execution. The layout design of a ­single‐use facility will be achieved by using specific design principles with focus on the process technology. Process closure, like open vs. closed processing, as well as different segregation principles, like physical, timely, or procedural segregation, have to be considered. One of the main targets to lower the manufacturing costs is lowering the required cleanroom area and it’s classification by evaluating the opportunities to perform the bioprocess unit operations in a functionally closed manner. Isolating a process from its environment mitigates a major risk of contamination by adventitious agents, thereby yielding a process that is  safer, more robust, more consistent, and where the  risk of product contamination is reduced significantly. Technological improvements on single‐use products have already resulted in more robust systems that can isolate bioprocess unit operations from the immediate environment. This helps in controlling physicochemical conditions within a process microenvironment which is more achievable than control of the macro‐environment surrounding that process. Closed processing represents the lowest risk option in terms of mitigating the risk of contamination from the environment for both aseptic and low bioburden operations and enables the realization of ballroom environments. Detailed risk assessment for process closure and environmental background for process unit operations will be the enabler for its realization. Full attention shall be given to the design and s­ election of the consumable materials as well as their i­ ntegration into the process flow. As far as possible, standardized bag and filter assemblies shall be used. These predesigned solutions will substantially simplify the supply chain in terms of complexity, safety, and speed. The wide application of single‐use technologies in all c­ ritical process steps and applications require quality attributes that are more challenging to achieve, such as a complex supply chain with a bigger reliance on supplier quality system, change control and business continuity, the biocompatibility of the contact layer of the bag ­ film,  understanding of leachables and extractables, and the container closure integrity. It is key for the ­supplier of single‐use technologies to mitigate these risks. Assurance of quality supply can, for example, be enhanced via control of the entire process from the ­resins to final product, which in turn will also lead to consistent extractable profiles. Optimization of the resin formula will prevent issues with biocompatibility of the film. Assurance of integrity can be improved by point‐of‐use testing but also with a combination of tests at the supplier.

14.6 ­Process Automation for Commercial Manufacturing Facilities The process automation system (see also Chapter 7) has to satisfy at least the minimum Good Manufacturing Practice (GMP) requirements for reporting, data management, user handling, audit trail, and visualization. However, in modern facilities, automated process batch management and S88‐compliant batch recipe‐control functionalities as well as plant‐wide visualization and electronic batch records are to be considered as state‐of‐ the‐art technologies within the manufacturing execution systems (MES). The seamless integration of process equipment and process skids are a fundamental need and the flexible approach for movable system integration as well as supporting functions for single‐use handling are the basis for running commercial manufacturing facilities at maximum capacity with the assurance of highest quality. In addition to the automation system, the landscape shall also include advanced process analytical software such as multivariate data analytics tools. The use of PAT enables developers to build quality into the product and operators to track the current process compared to the most optimal  –  this is known as the “Golden Batch Analysis.” The process automation concept can also be aligned with the facility automation concept for the environmental monitoring system (EMS) as well as the building monitoring system (BMS) and some integration to ­commercial enterprise resource planning (ERP) systems (Figure 14.5).

14.7 ­Intensified Upstream Processing The upstream process within the design study of an intensified process facility is based on operating six single‐use stirred tank reactors (2000 l) with an appropriate cell and product retention device. The production bioreactor setup with associated retention devices will withdraw and replace cell culture media at a rate of maximally 1 vessel volume per day. It has been assumed that each 2000 l bioreactor generates a titer of 10 g/l of product, when concentrated fed‐batch is used as the mode of operation, and will, therefore, produce 20 kg of mAb per batch. Preliminary experiments have demonstrated that this level of productivity is easily achievable with commercially available Chinese Hamster Ovary DG44 expression systems [6, 7]. These experiments have shown that intensified culture mimics with this cell line leads to a threefold increase in productivity over 12 days and a

187

WFI loop 80°C

Media preparation room 1

POC

WFI loop 80°C POC

CW-S

CW-S

CS CW-R

CS CW-R

1 2

WFI 20°C

1 2

Bio-Welder Bio-Sealer

WFI 20°C Sartocheck

Buffer

Buffer

BAG-1234

BAG-1234

xl (x l )

xl (x l )

Buffer MP9030-BAG-4180

AC-1000

xl (x l )

P-1234

P-1234

MP9030-P-4036

Powder Powder

Powder

MP9030-PB-4001

Alternative PB-1234

PB-1234 PAT-1234 BAG-1234 F-100

200 l (xxx l )

PAT-1234 BAG-1234

F-100 PAT-1234 BAG-1234

BAG-1234

xl (x l ) F-100

50 l (xxx l )

xl (x l )

500 l (xxx l )

PAT-1234 BAG-1234

PAT-1234 BAG-1234

BAG-1234

F-100

xl (x l )

2000 l (xxx l )

PAT-1234 BAG-1234

xl (x l )

M

M

AG-1234

M

AG-1234

MP9030-AG-4030

P-1234

MP9030-P-4035

P-1234

50 – 200 l Media mixing system

500 l Media mixing system

MP9010

2000 l Media mixing system

MP9020

MP9030

Media

Media

MP9030-PAT-4920 MP9030-BAG-4920

MP9030-PAT-4940 MP9030-BAG-4940

2000l (1380l )

2000l (2000l )

MP9030-F-4013

Media

Media

MP9030-PAT-4920 MP9030-BAG-4920

MP9030-PAT-4940 MP9030-BAG-4940

2000l (1380l )

2000l (2000l )

MP9030-F-4013

Media

Media

MP9030-PAT-4920 MP9030-BAG-4920

MP9030-PAT-4940 MP9030-BAG-4940

2000l (1380l )

2000l (2000l )

MP9030-F-4013

2000l (2000l ) MP9030-F-4014

2000l (2000l ) MP9030-F-4014

2000l (2000l ) MP9030-F-4014

Drain

Seed culture room

Cell culture room

Media

Media

Base

Media

5l (2 l )

5l (2 l )

PAT-1234 BAG-1234

PAT-1234 BAG-1234

200 l (60 l )

500 l (380 l )

BAG-1234

20 l (4/20 l )

BR1050-BAG-830

Media

Base

Antifoam

BAG-1234

BAG-1234

10 l (10 l )

10 l (10 l )

Base

Glucose

PAT-1234 BAG-1234

PAT-1234 BAG-1234

50l (30l )

100l (80l )

Antifoam BAG-1234

10 l (10 l )

Base

Glucose

PAT-1234 BAG-1234

PAT-1234 BAG-1234

50l (30l )

100l (80l )

Antifoam BAG-1234

10 l (10 l )

Base

Glucose

PAT-1234 BAG-1234

PAT-1234 BAG-1234

50l (30l )

100l (80l )

Antifoam BAG-1234

10 l (10 l )

BR1050-BAG-710 P-1234

P-1234

P-1234

P-1234

P-1234

P-1234

P-1234

P-1234

P-1234

P-1234

P-1234

M P-1234

P-1234

M

M

P-1234

BR1050-P-718

P-1234

STR-1234

P-1234

from BR1040

HydroSart ATF 30kDa 1 vvd

P-1234

STR-1234

P-1234

200 l (100 l )

500 l (500 l )

HydroSart ATF 30kDa 1 vvd F-1

BAG-1234

2000 l (2000 l )

STR-1234 BAG-1234

STR-1234

BAG-1234

2000 l (2000 l )

M P-1234

STR-1234 BAG-1234

P-1234

F-1

BAG-1234

100 l M

from BR1040

HydroSart ATF 30kDa 1 vvd

F-1 BR1050-P-838

2000 l (2000 l )

Drain

Drain

Drain

BW 20000 l

BW 20000 l

BW 20000 l

TITRE 10g/l

TITRE 10g/l

TITRE 10g/l

BAG-1234

1ml Vial

125ml (25ml ) BT-1234

1x (N-7)

1l (100ml )

2l (1000ml )

BT-1234

BT-1234

2x

1x

(N-6)

(N-5)

2000l

2000l

Biostat RM 50 Basic w/v 5 l / 25 l

1

(N-4/N-3)

(N-2)

RM 50 seed bioreactor

BR1010

2000l

M

BRM-1234

Shake flasks

(N-1)

200 l seed bioreactor

BR1020

P-1234

BR1030

1

P-1234

500 l seed bioreactor BR1040

1

P-1234

2000 l production bioreactor

P-1234

2000 l production bioreactor

BR1050

2000 l production bioreactor

BR2050

Bio-Welder Bio-Sealer

BR3050

Sartocheck

AC-1000

Media

Base

Media

5l (2 l )

5l (2 l )

Media

Media

PAT-1234 BAG-1234

PAT-1234 BAG-1234

200 l (60 l )

500 l (380 l )

BAG-1234

20 l (4/20 l )

BR1050-BAG-830

Base

Antifoam

BAG-1234

BAG-1234

10 l (10 l )

10 l (10 l )

Base

Glucose

PAT-1234 BAG-1234

PAT-1234 BAG-1234

50l (30l )

100l (80l )

Antifoam BAG-1234

10 l (10 l )

Base

Glucose

PAT-1234 BAG-1234

PAT-1234 BAG-1234

50l (30l )

100l (80l )

Antifoam BAG-1234

10 l (10 l )

Base

Glucose

PAT-1234 BAG-1234

PAT-1234 BAG-1234

50l (30l )

100l (80l )

Antifoam BAG-1234

10 l (10 l )

BR1050-BAG-710 P-1234

P-1234

P-1234

P-1234

P-1234

P-1234

P-1234

P-1234

P-1234

P-1234

P-1234

M P-1234

P-1234

M

M

P-1234

BR1050-P-718

P-1234 P-1234 STR-1234

from BR2040

HydroSart ATF 30kDa 1 vvd

P-1234

STR-1234

100 l M

M P-1234

P-1234

STR-1234 BAG-1234

200 l (100 l )

500 l (500 l )

P-1234

STR-1234

HydroSart ATF 30kDa 1 vvd F-1

F-1

BAG-1234

BAG-1234

BAG-1234

2000 l (2000 l )

2000 l (2000 l )

2000 l (2000 l )

STR-1234

BAG-1234

from BR2040

HydroSart ATF 30kDa 1 vvd

F-1 BR1050-P-838

Drain

Drain

Drain

BW 20000 l

BW 20000 l

BW 20000 l

TITRE 10g/l

TITRE 10g/l

TITRE 10g/l

BAG-1234

1ml Vial

125ml (25ml ) BT-1234

1x (N-7)

1l (100ml )

2l (1000ml )

BT-1234

BT-1234

2x

1x

(N-6)

(N-5)

2000l

2000l

Biostat RM 50 Basic w/v 5 l / 25 l

1

(N-4/N-3)

(N-2)

50 l seed bioreactor

BR2010

(N-1)

200 l seed bioreactor

BR2020

BR2030

P-1234

Media

Media

MP9030-PAT-4920 MP9030-BAG-4920

MP9030-PAT-4940 MP9030-BAG-4940

2000l (1380l )

2000l (2000l )

POC

CW-S

CS CW-R WFI 20°C

CS CW-R

1 2

1 2

Bio-Welder Bio-Sealer

WFI 20°C Sartocheck

Buffer

Buffer

BAG-1234

BAG-1234

xl (x l )

xl (x l )

Buffer MP9030-BAG-4180

P-1234

AC-1000

xl (x l )

P-1234

MP9030-P-4036

Powder Powder

Powder

MP9030-PB-4001

Alternative PB-1234

PB-1234 PAT-1234 BAG-1234

PAT-1234 BAG-1234

200 l (xxx l )

F-100

F-100 PAT-1234 BAG-1234

BAG-1234

xl (x l ) F-100

50 l (xxx l )

xl (x l )

xl (x l )

2000 l (xxx l )

PAT-1234 BAG-1234

xl (x l )

M

M

M

AG-1234

AG-1234

MP9030-AG-4030

P-1234

MP9030-P-4035

P-1234

50 – 200 l Media mixing system MP9100

PAT-1234 BAG-1234

BAG-1234

F-100

500 l (xxx l )

PAT-1234 BAG-1234

500 l Media mixing system MP9110

2000 l Media mixing system MP9120

Figure 14.4  Example of a process flow diagram for 6× 2000 l USP setup.

2000 l production bioreactor

BR5050

Media

Media

MP9030-PAT-4920 MP9030-BAG-4920

MP9030-PAT-4940 MP9030-BAG-4940

2000l (1380l )

2000l (2000l )

MP9030-F-4013

BR6050

Media

Media

MP9030-PAT-4920 MP9030-BAG-4920

MP9030-PAT-4940 MP9030-BAG-4940

2000l (1380l )

2000l (2000l )

MP9030-F-4013

2000l (2000l )

CW-S

P-1234

2000 l production bioreactor

BR4050

WFI loop 80°C

2000l (2000l ) MP9030-F-4014

1

P-1234

2000 l production bioreactor

BR2040

MP9030-F-4014

POC

1

P-1234

500 l seed bioreactor

MP9030-F-4013

WFI loop 80°C

Media preparation room 2

2000l

M

BRM-1234

Shake flasks

2000l (2000l ) MP9030-F-4014

14.8  Intensified Downstream Processing

Process data analytics L3

PAT

Energy management

Process data warehouse OEE

EMS

MES

Data capture (Historian, PI system...) SCADA

L2

BMS/EMS

DCS PLC L0/1

Utilities/process equipment facilities

Figure 14.5  ISA S95 structure: Four‐level concept for manufacturing facilities. Overall equipment effectiveness (OEE); distributed control system (DCS); supervisory control and data acquisition (SCADA); programmable logic controller (PLC).

fourfold increase over 15 days. Assuming 20 runs per bioreactor per year, the company will be able to schedule the six vessels so that they can run 120 batches per year. This will allow approximately 1800 kg of antibody to be synthesized each year. The above output is based on a concentrated fed‐batch mode as described (Table 14.1). The upstream manufacturing strategy is highly flexible and can be adjusted depending on the market demand and the titers achieved during development. It considers the following process options: ●●

●●

●●

Use of a high cell‐density cryobag as a start point of the batch; this way the process can be started at the N‐2 seed. The advantage this brings is that there will be no need to have a dedicated suite for expansion of the working cell bank. Moreover, due to the totally closed operation, any contamination risk is reduced to a minimum. N‐2 rocking motion bioreactor or N‐1 stirred tank reactor in perfusion; this allows to skip a seed bioreactor in case of the N‐2 rocking motion bioreactor in perfusion or seed the final bioreactor (N) at a higher cell density, reducing the length or maximizing the output of the step. In Figure 14.6, recent data from a 2 l rocking motion perfusion bioreactor demonstrates that it is possible to achieve highly viable cell densities (VCC) in this culture format [6]. Data like this show that there are several options to redesign and optimize the classical seed train used in the industry nowadays. Production bioreactor in perfusion mode; this enables operation in concentrated fed‐batch mode, maximizing the upstream titer up to at least 10 g/l in a standard 12 day process time.

A significant challenge faced by engineers designing facilities with intensified upstream processes is being able

120

100

100

80

80

60

60 40

40

VCC

20 0

20

Viability 0

1

2

3 4 Time (d)

5

6

Viability (%)

ERP VCC (106 cells ml–1)

L4

7

0

Figure 14.6  Cell density profile of a rocking motion bioreactor run in perfusion [6]. Source: Courtesy of Sartorius Stedim Biotech.

to provide the bioreactor suites with sufficient cell culture media. Supplying 6× 2000 l bioreactors at one bioreactor volume a day will require media production operations to supply 12 000 l of cell culture media every 24 hours. The facility concept has to consider the required quantities as well as the media distribution to the related production bioreactors (Figure 14.7). The concept considers adjacent placements of media hold vessels to each of the bioreactors as well as adjacent ­harvest technologies toward the center of the overall arrangement. To achieve the specified upstream productivity based on the preliminary cell culture experiments, we estimate that each bioreactor will generate cell densities around 35 million cells/ml, equivalent to 10% wet cell weight and 200 kg of biomass per bioreactor batch. Removing this quantity of biomass during bioreactor harvesting requires operators to transfer the contents of the bioreactor to a separate 2000 l single‐use mixing vessel before the addition of approximately 50 kg of body feed derived from diatomaceous earth. The contents of the mixing vessel are then clarified by body feed filtration using ultrapure diatomaceous earth as a filter aid. The system is completely single‐use, highly scalable, and produces very consistent results. The biomanufacturer will need 32 cassettes to clarify 200 kg of biomass. No additional flocculating agent is required and the cassettes do not need pre‐flushing. Membrane filters installed downstream of clarification cassettes ensure low bioburden and a high‐clarity supernatant.

14.8 ­Intensified Downstream Processing Our conceptual design for the intensified facility utilizes a  standard Protein A capture chromatography step. Performing this step in eight hours would require five cycles on a 77 l column with a dynamic binding capacity of 50 g/l. To improve the throughput of the downstream operations, eluates from individual protein A chromatography cycles

189

190

14  Design Considerations Towards an Intensified Single‐Use Facility

Figure 14.7  3D representation of the upstream suite from the intensified single‐use facility concept, showing 6× 2000 l single‐use bioreactors set up around a centered media distribution system and harvest technologies.

are treated separately at low pH in order to inactivate viruses. This step is automated using a S88‐compliant recipe‐based control virus inactivation system (Figure  14.8). Separate single‐use mixers are used for the downward and upward titrations of the eluates in this setup. Each treated eluate is then filtered into a 1500 l single‐use mixer to be pooled. Our design allows operators to perform a two‐ stage filtration step into this p ­ ooling tank because of the possibility that the concentrated protein solutions may precipitate during pH adjustment. To accommodate the different upstream production scenarios as described above, the facility is set up using a classical two‐step polishing procedure. The first step is a resin‐based cation‐exchange step in bind and elute mode. The eluate is directly loaded onto the next step, which uses a salt‐tolerant anion‐exchange membrane adsorber, in flow‐through mode. The membrane adsorbers are used in cassette format. This new format enables to adjust the required amount of membrane volume to the specific requirements of the process. In this way it provides the

Figure 14.8  S88‐compliant recipe‐based control system. Source: Courtesy of Sartorius Stedim Biotech.

­  References

flexibility needed in the intensified facility. The absence of hold steps not only increases throughput but also helps minimize the footprint of DSP suites. The eluate from the combined polishing step is virus filtered into a separate post‐virus purification room. A S88‐compliant recipe‐ based control viral reduction system performs the virus filtration continuously. Two specifically designed prefilters with a capacity of 5 kg/m2 are connected in parallel to protect two downstream hollow fiber nanofilters for the robust clearance of viral contaminants. For concentration and diafiltration of the antibody product, 14 m2 single‐use cassette units have been ­developed that allows the step to be completed in under four hours. The step is controlled by a S88‐compliant recipe‐based control single‐use crossflow filtration system with a 500 l recirculation mixing tank. The retentate is then formulated by the addition of excipients in an in other single‐use mixer. The volume of the product stream at this stage is approximately 230 l. The formulated product is sterile filtered into 15 storage containers which each have a volume of 16.6 l. This allows the formulated product to be frozen under controlled conditions using three cycles in an appropriate freeze and thaw system.

compared to the blockbusters of recent years. This calls for a change in manufacturing strategy. Recent advances in cell culturing technologies have made it possible to reach higher titers in upstream processes, through different types of manufacturing scenarios. This creates the opportunity to move away from classical 12 000 l stainless‐steel facilities into flexible facilities, set up around 6× 2000 l single‐use bioreactors feeding into a single DSP train. This chapter describes the considerations during the design phase of the intensified facility design concept. Several approaches can be taken upstream. One option is to intensify the N‐1 seed bioreactor, to shorten the seed train duration, and make it possible to inoculate the production bioreactor at a high cell density. Another option is to use a concentrated fed‐batch mode of operation in the production bioreactor. Both strategies will increase the upstream titer. Harvest and purification technologies considered for implementation in the platform process need to accommodate the variability in titer upstream. As analyzed, this facility can yield 500–2000  kg/yr, depending on the upstream manufacturing scenario at a COGs as low as 50 $/g.

14.9 ­Summary and Conclusions

­Acknowledgments

The development of next‐generation molecules such as bispecific antibodies, and biosimilars, will lead to products tailored to more specific patient groups. Therefore, these molecules will need to be produced in lower q ­ uantities,

The authors would like to thank everybody at Sartorius Stedim Biotech GmbH, who contributed to the contents of the chapter, especially the colleagues from the process modeling and engineering teams.

Nomenclature BMS CAPEX COGs DCS DSP EMS ERP GMP MES

Building monitoring system Capital expenditure Cost of goods Distributed control system Downstream processing Environmental monitoring system Enterprise resource planning system Good Manufacturing Practice Manufacturing execution systems

mAb NA OEE PAT PLC SCADA USP VCC VV

Monoclonal antibody Not available Overall equipment effectiveness Process analytical technologies Programmable logic controller Supervisory control and data acquisition Upstream processing Viable cell density Vessel volume

­References 1 BioPhorum Operations Group. (2017).

Biomanufacturing technology roadmap. http://www. biophorum.com/wpcontent/uploads/2017/07/ SupplyPartMgmnt.pdf (accessed 2 August 2018). 2 Hutchinson, N., Manzke, C., Monge, M., and Gupta, P. (2018). Could commercial manufacturing with single‐use

technologies increase agility? https://www.pharmpro.com/ article/2018/03/could‐commercial‐manufacturing‐single‐ use‐technologies‐increase‐agility (accessed 2 August 2018). 3 Bisschops M. (2015). Filling the void: the key element of continuous processing. BioProduction, Dublin, Ireland (14–15 October).

191

192

14  Design Considerations Towards an Intensified Single‐Use Facility

4 Singapore Economic Development Board. (2018). Press

release “Amgen unveils next generation biomanufacturing facility in Singapore.” https://www.singaporebusiness. com/2014/amgen‐unveils‐next‐generation‐ biomanufacturing‐facility‐in‐singapore.html?sub FormJan2018=visible (accessed 22 January 2018). Mothes, B., Pezzini, J., Schroeder‐Tittmann, K., and 5 Villain, L. (2016). Accelerated, seamless antibody purification. BioProcess Int. 14 (5): 34–58. Matuszczyk, J., Schulze, M., Janoschek, S., et al. 6 (2018). High cell density cell cultures (>100E6 c/

mL) in 2D bags with integrated filter for seed train process intensification. https://www.sartorius.com/ mediafile/20180313_BSP_2018_Lisboa_JM‐gz‐02%20 (002).pdf (accessed 2 August 2018). 7 Zijlstra, G., Carpio, M., and Hutchinson, N. (2018). Perfusion in automated single‐use Minibioreactors. BioProcess Int. 16 (6): Suppl. https://www.sartorius.com/ mediafile/Accelerating_Intensified_Bioprocesses_with_ High‐Throughput_Small‐Scale_Tools.pdf ): accessed 2 August 2018.

193

15 Single‐Use Technologies in Biopharmaceutical Manufacturing A 10‐Year Review of Trends and the Future Ronald A. Rader and Eric S. Langer BioPlan Associates, Inc., Rockville, MD, USA

15.1 ­Introduction Single‐use bioprocessing equipment has made considera­ ble progress in the past 10 years, and will see further devel­ opment and adoption in coming years. The single‐use equipment market has grown in recent decades from a few legacy products, such as plastic serum and media storage bags, tubing and filter membranes, with limited markets to the current situation where single‐use technologies repre­ sent the majority of non‐commercial (preclinical and clini­ cal scales) applications in bioprocessing. This includes a wide variety of products and novel technologies currently available, and in development. Our 15th Annual Report and Survey of Biopharmaceutical Manufacturing [1] pro­ vides a global analysis of the bioprocessing industry and shows that single‐use equipment is now dominating small‐ and mid‐scale bioprocessing and starting to graduate to adoption for larger scale commercial manufacturing. Single‐use systems (SUS) are now commonly used by both developers and contract manufacturing organizations (CMOs), and have advanced technologically and in terms of their adoption. With advancing technology, knowledge, adoption, and experience with single‐use‐based bio­ processing, future progress and market expansion are expected. These expansions and adaptations of the single‐ use system (SUS) technologies now implemented in main­ stream biologics production are also being seen in adjacent segments, including cell and gene therapy, and even in the production of cosmetics, and food derived from plant cell and tissue cultures, which are now being produced at ­manufacturing scales.

15.2 ­Background Single‐use equipment is generally composed of plastic parts that are gamma‐irradiation sterilized and used once (or reused for a single‐product manufacturing ­campaign)

and discarded. By now, there has been over a decade of combined industry experience covering the benefits of single‐use vs. fixed stainless steel, including concerning SUSs’ lower capital investment and operational costs and flexibility. SUS are generally recognized as enabling rapid setup of bioprocessing and progressive manufacture of multiple products at multiple scales in the same manu­ facturing areas. Our single‐use definition concentrates on the product as sold and used. Legacy SUS used for decades include silicone and other plastic tubing and simple filter mem­ branes. Over the past 10 years, single‐use product lines have expanded to include devices from basic storage bags to complex bioreactors. Now, essentially all, par­ ticularly, upstream bioprocessing can be done with SUS. Downstream processing using single‐use technologies, particularly many types of chromatography, lag behind; at present, it is difficult, expensive, and rare to imple­ ment fully single‐use downstream processing. Bioreactor sensors, probes, and related automation are another area where suitable single‐use technologies and products are lagging and generally not yet available. Single‐use micro­ bial manufacture, which is now a small niche compared to mammalian bioprocessing systems, has also resisted development of SUS, particularly bioreactors, with the very highly energetic mixing, higher temperatures, requirements for heating and cooling, and other aspects restricting use of the current plastics‐based equipment technology. The great majority of SUS is being used for mammalian cell culture. Part of the growing dominance of mammalian cell culture is that it is amenable to use of single‐use bioreactors. Bioreactors (see also Chapter 4) and mixers (see also Chapter 3) anchor upstream bioprocessing. In the early 2000s, just a few SUS, such as simple WAVE bag bioreac­ tors on rocker platforms and a few primitive fixed‐wall bioreactors with plastic liners, were available, with these generally limited in size, to a few 100 l. Now, a variety of

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

194

15  Single‐Use Technologies in Biopharmaceutical Manufacturing

single‐use bioreactors and mixers are available at ≤2000 l scale. Due to engineering limitations, with each 1000 l weighing about 2200 lbs, SUS above 2000 l are generally not practical or cost‐effective. Currently, 1000–2000 l bioreactors are on track to be become the industry stan­ dard for new product large‐ and much commercial‐scale manufacturing, with a large and growing number of companies offering an increasing range of product options. However, technological progress in other upstream areas has also been dramatic, including improved cell lines, expression systems, and culture media, such that expression yields have been doubling about every five years, with this expected to continue. Thus, more bioprocessing can increasingly be done using smaller or the same scale equipment. A decade ago, a facility producing ­several 100 kg/yr of monoclonal anti­ bodies would have required multiple ≥10 000 l stainless‐ steel bioreactors and other comparably scaled equipment, while today, the same amount can be manufactured with a few 500–1000 l bioreactors, including single‐use, oper­ ating continuously, in parallel or at multiple facilities, and at much lower cost. About a decade ago, there were significant unknowns and concerns about single‐use equipment polymer lea­ chates and extractables (L&E, see also Chapters 11 and 17) which threatened to hold back the adoption of SUS. This included concerns about related toxic contami­ nants and regulatory agency disapproval of products manufactured with extensive, lengthy exposure to ­plastics. Today, these plastics and equipment have much improved, with multilayer laminated plastic bags ­universally used in single‐use bioprocessing, along with knowledge about reduction and removal of many bad‐ actor leachates. L&E continue to be perceived as a ­significant issue, but progress is being made, and both end users and suppliers recognize that they must work to address safety issues, including keeping up with ­supply‐chain modifications and information needed to perform toxicology assessments.

15.3 ­Methods This review of trends in the bioprocessing SUS market over the past decade‐plus is based on data from the annual survey of bioprocessing professionals conducted by BioPlan Associates, Inc., now in its 15th year [1]. The recent survey included responses from 352 individuals, including 222 working for biopharmaceutical develop­ ers/manufacturers, and, in separate analyses we include responses from 130 staff working for bioprocessing suppliers or vendors. This survey is long including ­ ­concentration on SUS, with this recognized as having considerable impact on capacity, and interest to those in

the industry. In addition to quantitative survey data, this annual survey and report includes extensive data and analysis of bioprocessing and related trends. We also include data from other primary sources, as noted.

15.4 ­Results 15.4.1  Market (Facilities) Distribution From our research, we identify single‐use technologies as a key factor in expanding global capacity and access to technologies for biologics manufacturing. This is now impacting the growth of capacity in regions such as China, where SUS technologies are being implemented at new facilities. The Chinese domestic industry is expanding at a rate that exceeds the global average, and much of that “greenfield” growth is being done using SUS strategies. This is allowing greater flexibility, and lower capital costs and investment, which, in turn, reduces risks. Regarding the current growth in global capacity, we present Table  15.1 which includes current estimates of worldwide biopharmaceutical manufacturing capacity, in terms of cumulative bioreactor volumes. This is based on BioPlan Associates’ Top 1000 Global Biopharmaceutical Facilities Index which includes data on well over 1000 bioprocessing facilities. Those facilities with over 500 l capacity account for 99.5% of the total capacity world­ wide [2]. Total worldwide bioprocessing capacity is now ~17 million l. Table 15.1 presents some related top‐level capacity and facilities data. Table 15.1  Worldwide capacity and facilities summary data. Percent of total

Parameter

Value

Total worldwide capacity

16 900 000

100

Total culture/fermentation capacity

15 400 000 l

91

Mammalian capacity

10 470 000 l

62

Microbial capacity

4 930 000 l

29

Blood/plasma/other expression systems

1 500 000 l

9

Number of facilities, ≥1000 l cap.

821

53

Number of facilities, ≥500 l cap.

1 074

69.9

Number of facilities, with any capacity

1 500+

—

Total capacity at facilities with ≥500 l

17 428 237

99.5

15.4 Results

Regional use of SUS generally parallels overall bio­ processing capacity, although there is more preference and increased use of SUS in developed vs. certain devel­ oping countries. With commercial manufacturing domi­ nating worldwide total capacity and with pre‐commercial manufacturing done only intermittently and at smaller scales, commercial manufacturing, still predominantly stainless steel‐based, thoroughly dominates worldwide bioprocessing capacity. An estimated 10% of current worldwide total bioprocessing capacity or ≤1.7 million l capacity involves primarily SUS process lines. 15.4.2  Single‐Use Systems Market Estimates The many advantages offered by SUS have resulted in their capturing a majority of the small‐ and mid‐scale markets for bioreactors and other upstream equipment; that is, for scales up to 2000 l. In contrast, most of the big‐ticket equipment items downstream, particularly chromatography systems, remain dominated by fixed or other reusable columns and other equipment. Table  15.2 presents estimates of various bioprocess­ ing‐related markets. Note, SUS are used once and dis­ carded, while fixed stainless‐steel equipment is endlessly cleaned and reused. Capital investments, the infrastruc­ ture involved, are much greater with stainless steel, while the recurring replacement costs involved with SUS can be prohibitive at certain scales. The growing market for commercial manufacturing using SUS is a primary driver for the SUS market’s future growth. Other drivers increasing future SUS use include a large number of biosimilar and developing country‐based developer/manufacturer facilities coming online, with most adopting SUS for commercial manufacturing. Note, the bioprocessing supplies market is presumed to be fairly evenly divided between up‐ and downstream applications.

Growth in the bioprocessing supplies market generally tracks that of the overall biopharmaceutical market, with this rather steadily increasing about 12–15% annually, roughly nearly doubling about every five years [3]. As shown above, the overall bioprocessing supplies market, both research and commercial manufacturing, is cur­ rently about 8% of total biopharmaceutical sales. BioPlan generally estimates that commercial biopharmaceutical product manufacturing costs are about 4–6% of product revenue. Stainless‐steel facilities currently account for ~85% of the market in terms of revenue, with single‐use at ~15%; this is projected to change in 5–10 years to a split of about 70–75% stainless vs. 25–30% single use. This includes growth in the SUS market, with this increasing perhaps 300%; and single‐use market share growing from current to ~25–30%. These data reflect down­ stream SUS applications still remaining limited five years out, with some increased adoption of SUS con­ tinuous chromatography, and increased use of SUS for commercial manufacturing. These projections also reflect likely market growth from many new players entering ­bioprocessing, including in developing coun­ tries and biosimilar developers. The most dramatic growth in SUS markets will be for commercial‐scale Good Manufacturing Practice (GMP) manufacturing. The single‐use market for commercial applications is projected to grow to over $1 billion/yr in the next five years. This growth will be driven by prod­ ucts currently in development using SUS receiving approval and graduating to commercial‐scale manufac­ turing. Pre‐commercial (preclinical and clinical) manu­ facture is sporadic, spread out over years, but commercial manufacturing is done at much larger scales and gener­ ally continuously for many years, often well over a ­decade of product life. Equipment costs for commercial ­manufacturing are much greater than non‐commercial

Table 15.2  Estimates of selected bioprocessing systems/facilities current markets (annual revenue) and growth in the past and upcoming 5 years. Products (revenue/year)

2013 Market

2018 Market

2023 Market

Biopharmaceuticals

$175 billion

$275 billion

$507 billion

Recombinant proteins

$85 billion

$125+ billion

$300 billion

Bioprocess equipment, total market

$12 billion

$23 billion

$40 billion

●●

Bioprocess equipment, upstream

$6 billion

$11.5 billion

$20 billion

●●

Bioprocess equipment, downstream

$6 billion

$11.5 billion

$20 billion

●●

Stainless steel/non‐SUS

$10.8 billion

$19.5 billion

$29 billion

●●

$1.4 billion

$3.5 billion

$11 billion

⚪⚪

Single‐use (SUS) Non‐commercial use (small/mid‐scale)

$1.3 billion

$3 billion

$9.5 billion

⚪⚪

Commercial use (large scale/GMP)

$0.1 billion

$0.5 billion

$2.0 billion

195

196

15  Single‐Use Technologies in Biopharmaceutical Manufacturing

manufacturing (although much lower per unit manufac­ tured). New products will continue to mostly be mono­ clonal antibodies, which with their relatively low potency will generally continue to require manufacture of ≥100 kg or much more. This will increasingly involve parallel trains, continuous bioprocessing, and/or multiple facili­ ties worldwide anchored by 500–2000 l SUS bioreactors. But, even with this SUS growth, fixed stainless‐steel ­systems will still dominate the GMP/commercial and, thus, overall market five years out. 15.4.3  Market Trends and Perceptions The BioPlan annual survey and other sources confirm that among the facilities surveyed, mostly in the United States and Europe, average purchases of single‐use equipment are over $1 million/yr. Bioprocessing profes­ sionals realize that SUS has improved bioprocessing. Over two‐thirds (68.8%) of survey respondents this year cited SUS as providing “some” or “significant” improve­ ments in their bioprocessing within the past year. But end users recognize that they are often using first‐ generation‐type or other less than optimal or desired equipment. Over the past five years, our survey indicates that end users continue to express desires for improved SUS. As with most emerging technology areas, once new and better products and technologies are intro­ duced and become the norm, demand for further improvements continues, until a general level of satisfac­ tion is reached. In the bioprocessing industry, end users are apparently becoming concerned that innovation in single‐use bioprocessing supplies has been slow in com­ ing (e.g. the now decades‐old plastic bag‐in‐a‐vessel ­paradigm remains unchallenged). This is fully expected and normal in highly regulated environments, such as ­biologics, and demand for better products continues. In our recent 2017–2018 survey, we found: ●●

●●

●●

●● ●●

The average projected five‐year (2023) increase in mammalian manufacturing capacity at facilities is 32 and 24% for microbial facilities 34% of industry respondents desired improvements in even the most basic devices such as bags and connectors 35% wanted improvements in single‐use and disposable probes and sensors 41% wanted cheaper SUS equipment and 40% wanted more single‐use purification options.

Figure  15.1 shows single‐use products with the reported highest rate of use (at any stage/scale). The most commonly used products are mostly simpler devices, such as bags (see also Chapter 2) and connectors (see also Chapter 5). But, it is significant that over 70% cite use of major bioprocessing systems, including

­ ioreactors, mixers, tangential flow filtration (TFF), with b these facilities more likely fully dedicated to using SUS. The BioPlan survey covers all aspects of bioprocessing, yet the top desired types of new bioprocessing suppliers were nearly all associated with single‐use applications. This was the case in most prior years of this now 15 year‐ old annual survey. The trend for better single‐use prod­ ucts is exemplified by increased purchases and budgets for these devices, especially over the past five years, and by increased research and development (R&D) focus among suppliers and technology innovators in the field. In terms of bioreactors, the highest percentage of sur­ vey respondents, 70.2%, reported that they would specify fed‐batch SUS bioreactors for any new facilities at clini­ cal scale and 51.9% would specify SUS bioreactors for new commercial manufacturing facilities. Note, the majority still now longer favor stainless steel at either clinical or commercial scale. About half now expect to see fully SUS facilities in five years. As expected, much growth in SUS has occurred in recent years. Products seeing the most rapid growth rates in adoption from 2006 to 2018 (from when survey for these data started) include membrane adsorbers, bio­ reactors, and mixing systems with annual growth both in the 20% range, with most others product areas having 50% of respondents, associated with single‐use were: ●●

●● ●● ●●

“Breakage of bags and loss of production material” cited by 75.0%; “Leachables and extractables” (L&E) cited by 73.3% “High cost of disposables” 68.8%; and “Material incompatibility with process fluids” and “We do not want to become vendor‐dependent ­(single‐ source issues)” tied at 56.7%

15.5 Discussion

90.6%

Tubing for disposable applications

86.2%

Disposable filter cartridges

81.8%

Bags, empty Connectors, clamps

80.5%

Buffer containers

80.5%

Media bags, filled (finished with liquid media)

78.6%

Preassembled tubing sets, rigging kits, etc.

78.0%

Depth filters

77.4%

Bioreactors

76.7%

Sampling systems

76.1%

Tangential flow filtration devices

76.1%

Waste containers

76.1% 72.3%

Mixing systems 61.6%

Disposable chromatography devices

59.7%

Membrane adsorbers Perfusion devices

46.5%

Figure 15.1  Usage of disposables in biopharmaceutical manufacturing, at any stage of manufacture, applications in biopharmaceutical manufacturing: % using single‐use products (for SUS users). Source: Courtesy of BioPlan Associates, Rockville, USA.

As use of SUS increases, the concerns expressed over SUS costs can be expected to increase, particularly as more bio­ processing professionals only familiar with stainless‐steel equipment increasingly purchase SUS. However, from other survey research we find that the price continues to not be a primary concern among purchasers of SUS or other bioprocessing equipment. End users predominantly prefer to purchase the best equipment they can get, and are willing to pay more, as long as the premium is not excessive. In this context, while SUS products are costly, this is an R&D intensive segment, and suppliers tend to invest heavily in the development of next‐generation technologies.

15.5 ­Discussion SUS manufacturing will advance and continue to ­proliferate, with more diverse technologies and prod­ ucts becoming available. However, in terms of capacity

­(cumulative bioreactor volume), most of the current largest facilities, e.g. multiple ≥10 000 l bioreactor, are likely to remain in use. New capacity will be added at lower scales, with most new commercial manufactur­ ing facilities still using stainless steel. As such, stainless steel will continue to dominate bioprocessing in terms of capacity and number of large‐scale and commercial facilities. A number of individualized biologics and personal­ ized medicines are in active development, including patient‐specific cellular and gene therapies, cultured tissues and organs, and cancer and other therapeutic vaccines. With their one‐off nature and needs for ­sterility, these products can all be expected to be man­ ufactured using single‐use equipment. BioPlan has projected significant growth in cellular and gene ther­ apy capacity and facilities, despite a current shortage that will become an actual “capacity crunch” in com­ ing years.

197

198

15  Single‐Use Technologies in Biopharmaceutical Manufacturing

55.7%

Bioreactors

52.9%

Mixing systems Membrane adsorbers

46.8%

Perfusion devices

46.5% 35.0%

Tubing for disposable applications Connectors, clamps

32.9%

Tangential flow filtration devices

32.6%

Media bags, filled (finished with liquid media)

31.0% 29.3%

Waste containers Preassembled tubing sets, rigging kits, etc.

28.5% 27.3%

Buffer containers 20.9%

Depth filters

18.2%

Sampling systems

17.3%

Bags, empty

14.6%

Disposable chromatography devices Disposable filter cartridges

8.0%

Figure 15.2  12‐Year percentage‐point change in first‐usage of disposables, 2006–2018. Note: Not growth in sales, this is growth in application first usage within a facility. Source: Courtesy of BioPlan Associates, Rockville, USA.

We are already seeing another SUS‐related trend  – modular facilities (see also Chapter 1), with bioprocessing unit processes housed in connectable trailer‐like portable cleanrooms or isolator units; and whole manufacturing facilities being portable, able to be constructed, including cloned/copied, and operational in a matter of months or even weeks. With modular systems using SUS equipment and providing much the same advantages as single‐use, modular unit markets are expected to grow dramatically (from a current very low baseline). This includes “plug‐ and‐play” factories, with whole production lines and ­facilities fully clonable. Modular systems will be particu­ larly beneficial in GMP‐challenged d ­ eveloping countries. With many developing countries increasingly demanding

domestic biopharmaceutical manufacture, modular facili­ ties will become more ­ commonly used in developing countries in the next ­decade. China and India, with their growing middle‐class populations, and demands for state‐ of‐the‐art healthcare are regions where biologics and bio­ similars are likely to be considered for manufacture using modular facility strategies. We can also expect an increase in the number of biop­ harmaceutical manufacturers and products in coming years, with the average scale and size of new manufactur­ ing facilities and process lines decreasing because of ­factors such as process optimization, increasing expres­ sion yields, and simply more products splitting mar­ kets. Currently, the biosimilars pipeline includes ~1000

15.6 Conclusions

­ roducts with involvement of over 170 companies, includ­ p ing many new players worldwide [4]. Single‐use, particularly, upstream bioprocessing has and will remain dominated by the plastic‐bag‐in‐stainless‐ container paradigm, which has not changed since the first single‐use products’ introduction. The use of ­flexible bags in bioprocessing presents a number of problems, including their insertion, securing, sealing, ports, and L&E. Novel approaches eliminating bags seem ­possible, such as unitary (solid) self‐supporting, rigid, disposable containers (no bag), or even inert materials used to line plastic containers. Such approaches may better support microbial bioprocessing using single‐use bioreactors. However, with many vendors and large‐scale users hav­ ing invested heavily in bag technology, few alternatives can be expected to start to gain traction in the market in the near term. Changes in bags in coming years likely will include: ●● ●●

●●

Increased adoption of more inert, Lower and less toxic L&E, contact‐layer fluoropolymer plastics, such as Teflon (polytetrafluorethylene); and we might even see bags with inner stainless‐steel foil con­ tact layers, which could resolve many L&E concerns. Similarly, other polymers and materials will be increas­ ingly adapted for tubing and connectors.

Currently, no real design or performance standards, even voluntary, yet apply to most SUS products, but indus­ try committees and groups continue to work in these areas. Downstream, we can expect selective, limited increased adoption of single‐use equipment, particularly among the chromatography systems that dominate this market and continue to resist single‐use implementa­ tions or replacement by new single‐use technologies. Multiple companies are starting to offer custom pre­ packed chromatography columns, but these columns, so far, are not truly single‐use, rather are returned to their suppliers for recycling. Protein A resins will remain, by far, the dominant method for antibody capture, with this only cost‐effective if the resin is multiply recycled. New(er) membrane adsorption technologies, single‐use moving bed, TFF, countercurrent chromatography and other novel alternatives to conventional chromatography methods implemented as single‐use will start to see more significant adoption at mid‐ and larger scales in coming years (see also Chapter 9).

15.6 ­Conclusions Fears of capacity shortages were prevalent over a decade ago, as multiple blockbuster monoclonal antibodies approached marketing and commercial manufacturing

capacity was limited at the time. But, these capacity con­ cerns have subsided as a result of: ●● ●● ●●

Increased productivity (primarily increased titers), Building of many new facilities and process lines, and Increased access to more rapidly deployable SUS as the major factors in resolving these concerns.

With single‐use and modular systems quickly installa­ ble and replaceable, capacity crunches are increasingly unlikely in the future. In fact, with SUS, estimating cur­ rent and projected industry capacity becomes increas­ ingly difficult, subjective, and potentially less relevant. Besides inherent conservatism in this highly regulated industry, an issue restricting progress in expanding the use of SUS is vendors’ hesitancy and end‐users’ distress over incremental improvements and other changes in estab­ lished product lines. This is because any changes in bio­ processing products already in use, particularly, at GMP can require extensive validation studies by vendors and end users, more regulatory ­filings, and often costly L&E testing, new standard operating procedures, etc. Innovations in single‐use equipment may more likely originate from small companies or from new major corporate entrants, both supplier and developer companies, less wedded to ­current, increasingly aging technologies. New technolo­ gies in this industry are generally and ­easily implemented by totally new facilities and process lines, not retrofitting. SUS, other than legacy products, such as tubing and fil­ ters, which already dominate their markets, will see con­ tinued technological progress and increased adoption and use for biopharmaceutical manufacturing. The fastest growing segments of the single‐use equipment market will be upstream bioprocessing at large and commercial scales, as products now in development using SUS move up to commercial manufacturing. As demonstrated by the annual BioPlan survey and confirmed by many other sources, adoption of SUS, generally in place of fixed stainless‐steel systems, will continue. Modular systems may well be the next technology to experience such increased and rapid adoption over the next decade. As cel­ lular and gene therapies emerge through the development pipeline, we will likely see SUS technologies created and adapted explicitly for these personalized applications. Most of the growth in the use of SUS will be in devel­ oped countries, but developing countries will see much faster growth rates (starting at near‐zero baselines). For  example, many of the new bioprocessing facilities expected to be constructed in China to meet both its domestic demand and desires to become an exporter of biopharmaceuticals will be primarily SUS facilities [5]. The future for SUS technologies remains defined by the expanding global need for better, cheaper, and faster ­biologics production.

199

200

15  Single‐Use Technologies in Biopharmaceutical Manufacturing

Nomenclature CMO Contract manufacturing organization GMP Good Manufacturing Practice L&E Leachates and extractables

SUS SUS TFF

Single‐use system Single‐use systems Tangential flow filtration

­References 1 BioPlan Associates. (2018). 15th Annual Report and

Survey of Biopharmaceutical Manufacturing Capacity and Production. BioPlan Associates. ISBN 978‐1‐934106‐33‐4. BioPlan Associates. (2018). Top 1000 global 2 biopharmaceutical facilities index. www.top1000bio.com (accessed 20 October 2018). Rader, R.A. (2014/2015). Biopharmaceutical 3 manufacturing: historical and future trends in titers,

yields, and efficiency in commercial‐scale bioprocessing. BioProcessing J. 13 (4): 47–54. Rader, RA. (2017). Biosimilars/Biobetters pipeline 4 directory. www.biosimilarspipeline.com (accessed 11 April 2017). BioPlan Associates (2017). Directory of Top 60 5 Biopharmaceutical Manufacturers in China, 2e, 357. (also online database).

201

Part II Application Reports and Case Studies

203

16 Single‐Use Process Platforms for Responsive and Cost‐Effective Manufacturing Priyanka Gupta, Miriam Monge, Amelie Boulais, Nitin Chopra, and Nick Hutchinson Sartorius Stedim Biotech GmbH, Göttingen, Germany

16.1 ­Introduction The increasingly competitive nature of the market for biopharmaceuticals (such as monoclonal antibodies [mAbs], antibody–drug conjugates [ADCs], vaccines, and cell and gene therapeutics) is exacerbating the need for greater cost efficiency within the industry. This increase in competition is predictable as the market matures and new firms enter the sector in search of growth and profits. Firms that attempt to gain a competi­ tive advantage by gaining “first‐mover” status must get their products into the clinic and then onto the market as quickly as possible. Those that can do this are typically able to take and retain significant market share. Arguably, this has been the case since the industry’s inception. The advent of biosimilar products, however, has changed the nature of competition within the industry. Products that are highly similar to innovator molecules are being brought to the market with lower costs due to a reduced need for clinical studies. Biosimilar companies can man­ ufacture these molecules using the latest technologies and techniques allowing them to be marketed at a ­significantly lower price [1]. This means that companies with innovator candidates must pay much closer ­attention to their manufacturing costs than they have previously done. Leading biopharmaceutical companies have recog­ nized that in order to maximize the return on their research investments, they must expand into geographi­ cal markets that have been untapped until now. Firms that have managed to saturate markets in North America and Europe are turning their attention to large countries with emerging middle‐class populations, such as India and China, who increasingly expect the latest healthcare innovations. To address these markets they are develop­ ing complex manufacturing networks with in‐house and outsourced production capacity located around the world. Such strategies, though logical, carry significant

risk. Entering new markets can be a challenge, so pre­ dicting the correct size of production assets to meet local market needs is challenging at best. Biopharmaceutical companies that are planning new production facilities must take all of these considerations into account. New biomanufacturing capabilities must be able to produce a broader range of products than has been needed in the past. The need for this capability implies that the facilities must be intrinsically flexible. However, rising product titers and process intensification technologies mean new facilities do not necessarily need to be large [2]. A number of biopharmaceutical companies are con­ sidering scaling‐out their production assets rather than scaling‐up to very large volume manufacturing. This comes with the additional benefit of reduced risk because firms can distribute their production facilities close to local populations and are not reliant on a single produc­ tion hub that could be brought off‐line by a natural ­disaster, labor‐related dispute, or serious contamination event. While process development scientists are chal­ lenged with developing very efficient production pro­ cesses that will produce product with a low cost of goods (COGs) in the shortest possible time, new production facilities must be constructed in the shortest time possi­ ble. It is advantageous if firms can delay investment in new sites until as late as possible when they have the most evidence that the candidate is likely to be success­ ful. In addition, the facilities can be built at any selected site in the globe and the processes can be transferred between locations readily. End‐to‐end single‐use ­production platforms from vendors help to address these challenges. This chapter will describe how biomanufacturers can streamline early‐stage development and rapidly config­ ure a standard and well‐qualified manufacturing single‐ use process by adopting the platform approach. The tangible benefits of this approach will be explored using

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

204

16  Single‐Use Process Platforms for Responsive and Cost‐Effective Manufacturing

a standard Monoclonal Antibody process (as defined by the Biophorum Technology Roadmap team). The focus will be on speed to clinic and to market, flexibility, and scale while using BioSolve modeling to quantify those benefits for commercial manufacture.

16.2 ­Standardized Single‐Use Process Platforms for Biomanufacturing Single‐use platforms are available for a wide range of ­biological entities including mAbs, viral vectors, ADCs, and regenerative medicines. The benefit of implement­ ing a platform from a supplier is that a significant part of the process design has already been performed and, therefore, does not need to be purchased a second time. In addition, not having to perform a complete facility design from scratch saves valuable time and ensures bio­ manufacturers can construct new facilities as quickly as possible and reach the market first [3, 4]. Figure 16.1 shows a single‐use process platform for a viral vector product. Leading vendors such as Sartorius are able to supply single‐use technology to meet all of the unit operations in the process train. Very often, they will have existing data to verify the performance of their ­technologies within these applications. The order of unit operations need not be fixed within the platform and suppliers may adopt a “toolbox” approach to help bio­ manufacturers establish the best process platform for their biologic ­ candidate. Suppliers might incorporate non‐processing technologies such as online, at‐line, and off‐line analytics, such as the ViroCyt® Analytics technol­ ogy shown in Figure  16.1 which is used for virus quantification. Single‐use platforms are adaptable and biomanufac­ turers can tailor them to their individual needs. If engi­ neers identify a need for a change in the design of disposable flow paths, this is readily achievable. Single‐ use facilities require a lower capital investment than their multiuse stainless steel (SS) equivalents even if the costs of consumables are higher. The benefit to bio­ pharmaceutical companies is that less upfront cost is committed at a stage when the uncertainty surrounding the product’s success is still high. If a product is unsuc­ cessful during development, then firms that have planned for single‐use facilities will have wasted less capital. While experts within the industry generally regard that single‐use platforms are limited to bioreactors with a 2000 l volume, the increasing productivity of modern bioprocesses makes such platforms a viable alternative to SS even for commercial‐scale production [5].

The “off‐the‐shelf ” nature of most single‐use auto­ mated systems allows identical platforms to be imple­ mented at production locations around the world, reducing risk and enabling localized production. Some industry commentators have expressed concerns about the sustainability of single‐use facilities but research shows that they require around 30% less electricity, 60% less steel, 90% less water, and 95% less cleaning solutions during operation compared to SS plants.

16.3 ­Implementing Single‐Use Process Platforms It is clearly critical for the successful implementation of single‐use process platforms that a life‐cycle approach (see also Chapter  13) is adopted from the discovery phase, through the preclinical and clinical phases and into commercial manufacturing. Along with having the appropriate technologies for each phase, suppliers must provide the expertise needed that will support their cli­ ents as they progress through the various critical activi­ ties. For example, during the discovery phase, companies will need high‐throughput screening technologies that help them define their therapeutic target. In the preclini­ cal phase they will need small‐scale technologies that help them select the process and the analytics needed for product characterization. Vendors should be able to provide the necessary exper­ tise needed to support development and optimization studies based on their knowledge of how their equipment performs at both the small‐scale and the large‐scale. Some vendors are using process cost modeling tools to help guide process development studies. Biomanufacturers expect suppliers of single‐use platforms to have fully scal­ able portfolios that allow process characterization and pilot plant studies before scale‐up to commercial produc­ tion. Vendors should be able to support their clients with process validation services including extractable and leachables testing. A considerable part of the value derived from single‐ use platforms comes from the integration of different processing technologies across the manufacturing train and then into the automation architecture of the pro­ duction site. Vendors have teams of engineers that will ensure that this integration occurs smoothly without the biomanufacturer having to perform these activities in‐house. This model provides scope for considerable cost savings as the expertise for this integration lies with the equipment supplier and it allows resources at the biopharmaceutical company to focus on their core competencies of developing and manufacturing the biologic.

Upstream process - Viral vectors Single-use centrifugation

Multi-parallel bioreactors

Depth filtration

Process development tool

Single-use mixing systems

Single-use storage bags

Sterilizing grade filters

Cell bank Single-use bioreactors

Virus seed Single-use bioreactors Pre-filtration

Mycoplasma retentive filters

Cell culture media

Containers

Seed expansion

Virus propagation

Optional - Microcarrier removal - Cell lysis

Media preparation Sterile filtration 0.2 μm

Mycoplasma retention 0.1 μm

Medium, feeds

Virus retention 0.02 μm

Analytics

Virus counter

Figure 16.1  A single‐use process platform for viral vectors. Source: Courtesy of Sartorius Stedim Biotech.

Cell removal

Downstream process - Viral vectors Sterilizing-grade filters Bioburden reduction

Control and monitoring Downstream intermediate filtration Sterilizing-grade filters Bioburden reduction

Clarification

Bulk harvest Buffer preparation Control and monitoring

Container

Sterilizinggrade filter

Single-use mixing system

Crossflow cassette

Concentration

Membrane adsorber

Capture

Membrane adsorber

Polishing

Buffer

Analytics

Virus counter

Figure 16.1  (Continued)

Crossflow cassette

Buffer exchange

Sterilizing-grade filter

Sterile filtration

USP seed train

16.4  Economic Analysis Comparing Stainless Steel with Single‐Use Process Platforms Standard stainless-steel fed-batch

Standard single-use fed-batch

Sartorius (SSB)single-use fed-batch

N-4 to N-1

N-4 to N-1

N-4 to N-1

Seed (5l to 500l)

Seed (5l to 500l)

Seed (5l to 500l)

stainless steel

single-use

single-use

Production bioreactor (2000 L SS)

Production bioreactor (2000 L SU)

Production bioreactor (2000 L SU)

Centrifugation

Harvest depth filtration

Harvest depth filtration

DSP seed train

Harvest depth filtration

Protein A

Protein A

Virus inactivation

Virus inactivation

Cation exchange - BƐtE (mode)

Cation exchange - BƐtE (mode)

Anion exchange FT (mode) traditional resin

Anion exchange FT (mode) sartorius membrane adsorbers

Virus filtration

Virus filtration

Ultra filtration/dia filtration

Ultra filtration/dia filtration

Sterile filtration

Sterile filtration

Figure 16.2  Process flow diagrams for three different process platforms.

16.4 ­Economic Analysis Comparing Stainless Steel with Single‐Use Process Platforms Sartorius Stedim Biotech uses process cost‐modeling tools to understand better the impact of process design and technology choices on overall process throughput and COGs. We use these tools to determine whether a bioprocess will be commercially viable and achieve the target COGs. The BioSolve Process software from Biopharm Services consists of datasets, unit operation libraries, and platform processes that include a cost breakdown for individual equipment, consumables, and materials. In addition, it allows analysis of multiple pro­ duction scenarios while comparing the impact of tech­ nological or unit operation changes on the COGs. These tools help companies identify the most optimized process flows. With this information, managers can direct scientists to invest their time and effort into optimizing these unit operations and deliver the best overall outcome. Companies can gain valuable insights from process and cost modeling software when considering the design of new biomanufacturing facilities. Process modeling with BioSolve Process allows to assess different production

paradigms during the conceptual phase of facility design. To demonstrate the economic value of single‐use process platforms, we used BioSolve Process to model three differ­ ent production scenarios. Each of the platforms modeled consisted of a single 2000 l production bioreactor in which a CHO cell line expressed a mAb biopharmaceutical to a titer of 3 g/l. The facility needs to produce 20 batches of product each year to meet the expected demand. The three different process platforms that were ­modeled are shown in Figure 16.2. The base case was a standard SS platform that incorporates a centrifuge for harvest and resin chromatography in the downstream process. An alternative scenario was modeled in which all unit operations in USP are single‐use while DSP is more or less the same apart from introduction of mem­ brane chromatography. The centrifuge was eliminated in favor of additional depth filtration area. In the final sce­ nario, a single‐use platform process (Sartorius’ single‐ use process platform for mAbs) was analyzed that replaces traditional ion‐exchange chromatography res­ ins with membrane adsorber technology. Figure 16.3a shows the throughput of product that each platform can deliver each year. The single‐use ­platforms deliver 91 kg while the SS platform provides 87 kg. This is

207

208

16  Single‐Use Process Platforms for Responsive and Cost‐Effective Manufacturing

because of yields’ losses over the additional centrifuga­ tion step. The SS process gives the most expensive prod­ uct at €102/g as shown in Figure  16.3b. The single‐use platforms give significantly lower cost product at approx­ imately €70/g. The Sartorius platform that replaces ion‐ exchange columns with membrane adsorbers gives a marginally lower cost product as the column chromatog­ raphy steps have higher capital and operating costs. Figure 16.4 allows a more detailed analysis for the dif­ ferences in the cost per gram of the product from the SS and single‐use platforms. It can be seen that the capital charge for the SS platform is significantly higher than for either of the two single‐use platforms. This capital charge includes not only the cost of the SS processing equip­ ment but also the cost of the additional utilities such as steam, water for injection, and cleaning solutions needed to run a SS plant. Furthermore, this capital charge incor­ porates the cost of all activities needed to install and commission successfully the equipment into the facility. Although the cost of consumables is higher when running single‐use platforms, it is more than offset by the higher capital charge. Figure 16.4 also shows that a ­single‐use facil­ ity requires less labor providing an additional saving. Process modeling with BioSolve Process allows devel­ opment scientists and process engineers to gain insights to which steps in the process are having the greater impact on COGs at different phases of the process development. Figure 16.5 shows which part of the process has benefited most from the switch to single‐use technologies. The per­ centage reduction in COGs is around 42–43% during upstream processing (USP) with single‐use technologies while the reduction is only around 24–28% during down­ stream processing (DSP) since only centrifuge step was eliminated in SU option and Anion Exchnage resin step was replaced by SU membrane chromatography step. Thus the benefit of switching to single‐use is more clearly observed in USP rather than DSP in this case study.

(a) Throughput kg/yr

93 91 89 87 85 83 81 79 77 75

Std Std SSB SS_mAbs_2000l SU_mAbs_2000l SU_mAbs_2000l

(b) Cost per gram EUR/g 120 100 80 60 40 20 0

Std SS_mABs_2000l

Std SU_mABs_200l

SSB SU_mAbs_2000l

Figure 16.3  Cost modeling data to show that single‐use platforms are capable of delivering higher throughputs (a) at lower costs (b) than stainless‐steel processes.

COG’s overall distribution 9 000 000 8 000 000 Std SS_mAbs_2000l Std SU_mAbs_2000l SSB SU_mAbs_2000l

7 000 000 6 000 000 5 000 000 4 000 000 3 000 000 2 000 000 1 000 000 0

Annual cost (EUR) Sum (Col. G to K)

Capital charge

Materials

Consumables

Labor

Figure 16.4  Breakdown of annual costs for a stainless‐steel mAb process and two single‐use mAb platforms.

Others

16.5  Summary and Conclusions USP- Euro per gram 80 70 60 50 40 30 20 10 0

DSP- Euro per gram

­ embrane adsorber during DSP, which reduces both the m initial capital outlay and the running costs of the facility.

28

24

16.5 ­Summary and Conclusions 42

43

91.1 Std SU_mAbs_2000l

91.1 SSB SU_mAbs_2000l

1 86.6 Std SS_mAbs_2000l

Figure 16.5  Reduction in COGs attributed to USP and DSP.

Finally, Figure 16.6 shows the net present value (NPV) of the three platforms. The single‐use platforms have the highest NPVs. The initial negative cash flow of a SS facil­ ity is extremely high relative to a single‐use platform even if the subsequent years’ positive cash flows are also higher due to the lower consumable costs. Importantly, the reduction in time it takes to construct a facility incor­ porating a single‐use production platform relative to SS plant means that the positive cash flows are realized sooner and are discounted less. The Sartorius single‐ use  platform gives the highest NPV because the use of

Single‐use process platforms allow biomanufacturers to address the challenges they face as the industry develops. They require less time and capital investment to install because they do not require the supporting utility infra­ structure associated with SS platforms. They are flexible and can be readily adapted as new information becomes available about the biological product and its associated manufacturing process. Vendors integrate the appropriate single‐use bioprocess­ ing technologies with one another and into the facility architecture so that biomanufacturers need not concern themselves with the connectivity of different systems and solutions. They should support companies with the imple­ mentation of process platforms during process develop­ ment, optimization, characterization, and qualification. Single‐use platforms allow biopharmaceutical companies to install new and highly efficient production capacity rapid­ly and with less upfront cost. This allows them to reach the market in the shortest possible time and with the lowest COGs for maximum competitive advantage.

Millons

NPC vs Throughput kg/yr 200 150 100 50 0 –50 0 –100 –150 –200 –250 –300 –350 –400 –450 –500 –550

NPV/NPC SU 1 Bioreactor facility

200

400

NPV/NPC SS 1 Bioreactor facility

600

800

1000

Figure 16.6  NPV of three mAb platform processes.

Nomenclature ADCs COGs DSP mAbs

Antibody–drug conjugates Cost of goods Downstream processing Monoclonal antibodies

NPV Net present value SS Stainless steel USP Upstream processing

1200

209

210

16  Single‐Use Process Platforms for Responsive and Cost‐Effective Manufacturing

­References 1 Blackstone, E.A. and Fuhr Joseph, P. (2013). The

economics of biosimilars. Am. Health Drugs Benefits 6 (8): 469–478. 2 Zijlstra, G. and Gupta, P. (2017). Moving toward continuous bioprocessing. https://www.genengnews. com/gen‐articles/supplement‐moving‐toward‐continuous‐ bioprocessing/6147 (accessed 14 October 2018). Boulais, A. (2018). Single‐use platforms accelerate viral 3 vaccine development and manufacturing. BioProcess Int. 16 (9): 7–11.

4 Monge, M. (2017). Single‐use production platforms for

biomanufacturing. BioProcess Int. 15 (11): 10–12.

5 Jones, S. (2017). BioPhorum operations group,

technology roadmapping, part 2: efficiency, modularity, and flexibility as hallmarks for future key technologies. https://bioprocessintl.com/business/economics/ biophorum‐operations‐group‐technology‐ roadmapping‐part‐2‐efficiency‐modularity‐flexibility‐ hallmarks‐future‐key‐technologies (accessed 14 October 2018).

211

17 Considerations on Performing Quality Risk Analysis for Production Processes with Single‐Use Systems Ina Pahl1, Armin Hauk1, Lydia Schosser2, and Sonja von Orlikowski2 1 

Sartorius Stedim Biotech GmbH, Göttingen, Germany Boehringer Ingelheim Pharma GmbH & Co. KG, Ingelheim, Germany

2 

17.1 ­Introduction The implementation of single‐use technology (SUT) into biopharmaceutical manufacturing processes has brought advantages to the industry (see also Chapter  1) [1, 2]. Nevertheless, new requirements for the qualification of materials of single‐use (SU) components by the European Pharmacopeia (EP) and United States Pharmacopeia (USP) [3, 4, 5] came up (compare also Chapter 11). For the evaluation of extractables and leachables (E/L), guidelines are available such as from European Medicines Agency (EMA) [6, 7] and Food and Drug Administration [8–11] which are originally intended for container ­closure systems (CCS). Additionally, USP draft [12] and USP draft [13] are currently under ­discussion covering SU components used in biopharma­ ceutical manufacturing. Even employed since over three decades, SUT faces limitations such as finite scale, restricted diversity of options, and lack of standardiza­ tion and lack of regulation of the quality of material used [1, 2]. A key concern for the implementation of SUT is the risk associated with visible and nonvisible particles and leachables related to SU components, to contami­ nate the final product and its intermediates [2]. Suppliers of SUT provide basic technical information for SU components within the validation/qualification docu­ ­ mentation. It can include data and results about the physical/mechanical strength of materials, functional testing, integrity and sterility of SUS, sterilizing meth­ ods, aging, endotoxins, biological reactivity, particulate matter, chemical compatibility, and extractables [1, 14]. A reasonable risk assessment concludes from extractable data to potential leachable data. Especially, during manufacturing of biologics, a ­consistent quality of SU components is required because the mechanisms by which the characteristics of the SU components affect the quality of biologics are not

c­ompletely understood [14]. For SU systems, critical quality attributes such as residual impurities including extractables, visible and nonvisible particulates, and endotoxin can be considered. In this context, suppliers of SUT need to have a rigorous quality management system including strong cooperation with suppliers, robust pro­ cesses with established materials and methods, test ­procedures, quality certificates, a focus on continuous improvement and innovation in processes and products, and effective communication with customers. In general, risk management for SU equipment starts with evaluation of potential risks related to the material, the process design, and the product [15, 16]. Table 17.1 lists some typical risks to SUS; however, this book ­chapter will concentrate on risk‐based approaches for leachables. It is distinguished between a common and holistic risk assessment, and the concept “Fate of Leachables” is introduced.

17.2 ­Quality Risk Assessment For quality risk management, the biopharmaceutical industry needs to understand the process dependencies and has to develop a risk‐mitigation strategy. Also, the industry has to communicate and interact with supplier and regulatory agencies on a trustful basis to identify the risks which affect safety of the patient, quality, and ensure supply of the final product [14, 16, 29]. Risk‐control methods are to be established which consider severity, probability, and detectability of deviations in process parameters [14]. In detail, Merseburger et al. [15] gave a comprehensive overview about the regulatory back­ ground for medicinal products. Key methods for risk assessment and hazards in SU manufacturing are described. The quality risk assessment of a SU device used in the production of final product is presented [15].

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

212

17  Considerations on Performing Quality Risk Analysis for Production Processes with Single‐Use Systems

Table 17.1  Typical risks related to SUS.

Material qualification

Process qualification

patient safety [1, 14, 30]. Several recommendations exist for E/L testing [5, 10, 12, 32–35] describing the fac­ tors  influencing the outcome of an E/L study. Recommendations for E/L risk assessment [13, 16, 31, 33] include risk factors such as propensity of contact fluid, contact time, contact temperature, surface to vol­ ume (S/V) ratio, daily dose plus route of administration, and proximity to the patient.

Description of risk

References

Biocompatibility tests

USP , USP [17] ISO 10993‐5 [18]

Additives

USP 661 [4], EP Chapter 3.1 [3]

Extractables

Vendor

Allergens (e.g. Latex)

n.a.

17.3 ­Terminology and Features

BSE/TSE

EMA/410/01 [19] EP 5.2.8 [20]

Particles

USP [21]

Endotoxins

USP [22]

Leachables

CFR 21 211.65 (a) [23] Eudralex Vol. 4 [24]

Mechanical stability

Various references given in Merseburger et al. [15], Lopes [1]

Gas transmission rate (for bags)

ASTM D3985 [24]

Sterility

USP [25] EP 2.6.1. [26]

Container closure integrity

ASTM D4991‐94 [27]

Filter integrity test (e.g. minimum bubble point)

PDA Report no. 26 [28]

Quality risk assessment according to International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use ICH Q9 is a process which defines responsibilities during risk management, initiation of risk management, risk assessment, risk con­ trol, risk communication, and risk review to e­ nable risk‐ based decisions concerning the quality of the final product (medicinal products) over the product life cycle. The risk evaluation should be based on scientific knowl­ edge using recognized management tools such as failure mode effects analysis (FMEA), hazard analysis and criti­ cal control points (HACCP), fault tree analysis, etc. and should reflect the complexity and criticality of the issue [29]. The outcome of the risk assessment can either result in a quantitative or qualitative description of a risk (Box 17.1).

Interaction with final product and component (e.g. adsorption of substances)

n.a.

ASTM, American Society for Testing and Materials; BSE, bovine spongiform encephalopathy; CFR, Code of Federal Regulations; ISO, International Organization for Standardization; n.a., not applicable; PDA, Parenteral Drug Association; TSE, transmissible spongiform encephalopathy.

Further information about the regulatory environment for SU equipment is given in Chapter  11 of this textbook. In short, before implementation of SUT, a risk‐assess­ ment approach is used to validate and qualify the SU components [1, 16, 30]. Exemplarily, as part of the pro­ cess validation using SUT, extractable compounds have to be evaluated and assessed because they may migrate as potential leachables from the polymeric materials into the production process at various stages [31]. They can either negatively influence the production process ­performance such as inhibition of cell growth or can contribute to process‐related drug impurities, which may potentially have an impact on product quality and/or

17.4 ­Current Industrial Approach for Leachable Assessment in Biopharmaceutical Processes As discussed extensively in Chapter 11, an assessment of potential leachables is today strongly focused on extract­ ables studies [6, 36] of individual SU components obtained under worst‐case or simulation conditions. Cumulative “worst‐case” models are used to estimate SUS leachables throughout a process step or the entire Box 17.1  Terminology Risk: The combination of the probability of occurrence of harm and the severity of that harm. Harm: Damage to health, including the damage that can occur from loss of product quality or availability. Hazard: The potential source of harm. Harm: Damage to health, including the damage that can occur from loss of product quality or availability. Severity: A measure of the possible consequences of a hazard.

17.4  Current Industrial Approach for Leachable Assessment in Biopharmaceutical Processes

process (Figure 17.1). From this approach, a high n ­ umber and high concentration of leachables in the final prod­ ucts can be expected. But, observations from real leach­ ables studies stand in contradiction to the above given expectation: process leachables do not contribute rele­ vantly to final drug impurities [36, 37]. In the biopharmaceutical industry, leachables can be assessed via a risk‐based approach. Extractables data available from the SU supplier play an important role dur­ ing this assessment. So, conditions of extractable experi­ ments have to be carefully evaluated if they represent conditions which are relevant for the process. As most of the biopharmaceutical processes are performed in aqueous solutions, extractables insoluble in water will ­ most probably not migrate into the product solution. Nevertheless, the draft of USP [12] describes extraction studies with 50% ethanol as a model for highly organic solutions or solutions containing surface‐active ingredients (e.g. typical excipients as polysorbate), the rec­ ommended standard approach of the BioprocessPhorum Operations Group includes 1% polysorbate 80 [35]. These model solutions may result in overestimation of potential leachables from extractable data. In a common risk‐based approach, evaluation of E/L can be divided into two steps. In a first step, each ­material is assessed during introduction in Good Manufacturing Practice environment (“material qualification”). The Step 1: Material qualification

Extractable risk assessment

Low risk

Step 2: Process specific evaluation (e.g. process validation)

No actions required

High risk

Relevant extractable data from supplier available

Leachable risk assessment of product-contacted surface materials

Low risk

No actions required

High risk

Yes

No

Perform extractable study

f­ ollowing points have to be considered for each material: process temperature, contact time of solution with mate­ rial, solvent (e.g. aqueous solution or organic solvents), surface area of the material, and the proximity to the final product. Based on these parameters, a risk priority number can be calculated resulting in a classification of material risk. Depending on the material risk, data for extractables has to be available or not (see Figure 17.1) [12, 13]. In a second step during process validation, a list of all product‐contacting material is compiled and a risk assessment is performed. The following parameters should be taken into consideration: process temperature, contact time of solution with material, solvent (e.g. aque­ ous solution or organic solvents), surface area of the material, proximity to the final product, pretreatments steps (e.g. pre‐flushing of material will reduce risk of leachables, additional sterilization may increase risk of leachables), and S/V ratio. The calculated risk priority number can help to focus on materials with higher prod­ uct risk for leachables. Typically, these are materials in downstream manufacturing process (e.g. after final depletion step such as tangential flow filtration [TFF]) and materials with large contact area and/or long contact times. For these materials, a worst‐case calculation of potential leachables based on the application regime (dose and route of administration) and relevant e­ xtractable

Paper-based assessment of extractables (potential risk)

Relevant extractable data available: paperbased evaluation in context to process sufficient

Yes

No

Perform leachable study

Figure 17.1  Exemplary scheme of evaluation of E/L information in the biopharma industry.

Evaluation of process specific patient risk

213

214

17  Considerations on Performing Quality Risk Analysis for Production Processes with Single‐Use Systems

data can be performed and toxicologically assessed. If a patient risk cannot be excluded, a leachable study for the final product may be necessary. From current experi­ ences, results for leachables studies are far below ­theoretical calculated values from extractable data.

17.5 ­Holistic Approach to Predict Leachables for Quality Risk Assessment Besides common risk assessments based on qualitative parameters (FMEA, HAACP, etc.), one can imagine that a quality risk assessment can also be set up on a quantita­ tive and holistic approach as an alternative. Such a holis­ tic approach predicting leachables within a generic up‐ and down‐stream process is based on mathematical model calculations, which describe all sources of leach­ ables and the role of sinks of leachables in a dynamically operated biopharmaceutical process. Several steps reduc­ ing potential leachables during downstream processing are given in Table 17.2. To obtain relevant information for the quality risk assessment of SUT, it is necessary to access quantitative predictions of potential leachables which consider well‐describable physical and chemical properties of the chemical entity and of the actual pro­ cessing steps during manufacture of a final product. The extraction or migration represents typically the source of leachables from SU components used in dynamic biopharmaceutical processes. The concentra­ tion of leachables can be described based on Fick’s second law of diffusion and mass transfer (“flux”) ­ through the polymer–solvent interface [44–46]. Mathematical solutions for the diffusion equation can be

found in Crank [47]. With this model, the kinetics and the final equilibrium concentration of leachables can be predicted [48–50]. Adsorption and desorption are reversible processes and significantly influence the levels of leachables in any biopharmaceutical manufacturing, e.g. in downstream filtration, separation, and purifica­ tion [37]. Process steps which remove the adsorbent out of the process stream can be regarded as final sinks for leachables. A downstream‐manufacturing process can be divided into compartments which are linked to specific process steps with defined SU components (compare Chapter 1). In each compartment sources, distribution and sinks of leachables can be calculated based on an underlying physical–chemical mechanism. Strict mass balance ­conditions need to be applied and the exchange between compartments or the discharge was modeled with the flow of the liquid phases [39, 40]. Figure  17.2 shows a dynamic box‐model to calculate the leachables load along the entire process. Table 17.3 shows the results of a model calculation for the concentration of two hypothetic leachable com­ pounds A and B based on the box‐model given in Figure 17.2. The physical–chemical parameters of leach­ able A are described as typical for an additive degrada­ tion product similar to e.g. a di‐tert‐butylphenol isomer and for leachable B similar to caprolactam. It is assumed that both compounds are present in the materials of the SU components. After the mixing process and storage for 24 hours, a migration out of the contact material during medium preparation is observed. For leachables A and B, a rising concentration in the medium in the bioreactor over

Table 17.2  Description of process steps as sources and sinks of leachables in biopharmaceutical processes.

Process

SU component

Objective of SU component

Physical effect

Harvest

Depth filtration Centrifugation

Cell removal, Clarification

Adsorption, Phase separation

Leachables of a bioreactor are adsorbed on host cell surfaces and cell debris [38]

Sterile filtration

Membrane filter

Particulate removal

Migration, Adsorption

Typical substance‐specific scavenger capacities for leachables in the range of some μg/cm2 of nominal membrane surface [37–39]

Purification

Membrane adsorber

Particulate removal, Concentration

Migration, Adsorption

Typical substance‐specific scavenger capacities were determined to be in the range of some μg/cm2 of membrane– adsorbent bed volume [38, 39]

Diafiltration

CrossFlow Cassette

Clearance, Particulate removal

Migration, Adsorption

Removal of leachables with UF/DF from the retentate [40]

Affinity chromatography

Column Flow kits

Clearance

Binding

Affinity chromatography is a highly selective binding process of a target molecule, all unbound substances can be washed out, so removal of leachables can be expected [41–43]

Scavenger effect

17.6  Summary and Conclusions Medium prep.: + mleach from raw materials + mleach by diffusion from contact materials

Bioreactor:

Harvest; depth filtration:

+ mleach increase by diffusion from contact materials during operation; leachables are adsorbed on biomass

+ mleach increase by diffusion from contact materials

Cell debris: – biomass

– mleach adsorbed on biomass

Different filtration steps: Purification steps: – removal of adsorbed mleach

Waste:

UF/DF–Filtration:

– permeate

– removal of mleach with permeate

+ mleach by diffusion from contact materials during transfer and filtration – removal of scavenged mleach

Handling after processing: + mleach by diffusion from contact materials during storage

In each compartment (box) the Σleachables are calculated with phys.-chem. methods: •

Diffusion-partitioning from contact materials

•

Adsorption on materials (biomass, membranes, etc.)

•

Dilution and/or concentration e.g. in crossflow steps

DP or DS in its CCS: + mleach by diffusion during storage

Figure 17.2  Elements of a dynamic box‐model including subsequent process steps in biopharmaceutical processing. DP, drug product; DS, drug substance. Source: Courtesy of Sartorius Stedim Biotech, Göttingen, Germany.

21 days is calculated. As leachable A shows a tendency to be adsorbed by the biomass, it is removed from the process by removal of the cell debris. This effect is much weaker for leachable B, as it is good water soluble with a lower tendency to be adsorbed onto the biomass. During subsequent handling and storage of the process fluid over 24 hours, the amount of both leachables is increasing. The filtration processes release compounds A and B by migration and remove them by adsorption onto the filtration membranes. The following ultrafil­ tration/diafiltration (UF/DF) step with TFF removes leachables A and B quite efficiently. Compound A has a lower z‐value and therefore a higher tendency to remain in the retentate. Leachable B with a z‐value of 0.7 is more efficiently removed from the process with the permeate stream. Although, both leachables A and B are present in the raw material and are migrating constantly out of the dif­ ferent contact materials, the combination of different downstream processing steps show a significant reduc­ tion of the leachables load down to very low levels in the final product. After processing, any storage of the prod­ uct, either in intermediate‐storage containers or drug‐ packaging systems can lead to rising concentration of leachables.

17.6 ­Summary and Conclusions SU technology becomes a key element in biopharma­ ceutical manufacturing. To evaluate the risk associated with leachables influencing the quality, efficacy, and patient safety, a current industrial approach for leach­ able assessment in biopharmaceutical processes applies the ­risk‐based approach. It starts with the eval­ uation of extractable data during material qualifica­ tion. Subsequently, all product‐contacting materials are assessed for the risk to release leachables into the final product. If a patient risk cannot be excluded, a leachable study may be necessary. The introduced concept “Fate of Leachables” considers the complete biopharmaceutical process and allows the calculation of the quantity of leachables in the final prod­ uct which can be used directly for the quality risk assess­ ment of the SU components used in the process. The model calculation confirms the intended purpose of a downstream process to purify the product by removal of undesired process impurities such as potential leacha­ bles. Additionally, the calculation demonstrates that the “proximity” to patient concept which is commonly used in risk assessments is reasonable and can be supported by the proposed model calculations.

215

­  References

Nomenclature ASTM American Society for Testing and Materials BSE Bovine spongiform encephalopathy CCS Container closure systems CFR Code of Federal Regulations D Diffusion constant DF Diafiltration DP Drug product DS Drug substance E/L Extractables/leachables EMA European Medicines Agency EP European Pharmacopeia FMEA Failure mode effects analysis HACCP Hazard analysis and critical control points ICH International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use ISO International Organization for Standardization KD‐bio Partition coefficient between biomass and liquid phase

Kapfilter Specific capacity of filters and purification devices KP/L Partition coefficient between plastic and liquid phase n.a. not applicable PDA Parenteral Drug Association SU Single‐use SUS Single‐use systems SUT Single‐use technology S/V Surface to volume TFF Tangential flow filtration TSE Transmissible spongiform encephalopathy UF Ultrafiltration USP United States Pharmacopeia z‐value Removal rate of compounds in a typical UF/ DF process; calculated according to equation: C C 0 e z Vadd /V0 [40] l

l

­References 1 Lopes, A.G. (2015). Single‐use in the biopharmaceutical

2

3 4

5

6

7

8

9

industry: a review of current technology impact, challenges and limitations. Food and Bioproducts Processing 93: 98–114. Abhinav, A.S. and Gottschalk, U. (2013). Single‐use disposable technologies for biopharmaceutical manufacturing. Trends Biotechnol. 31 (3): 147–154. European Pharmacopoeia (2016). 9e (1), 389–420. United States Pharmacopoeia. (2017). USP Plastic Packaging Systems and Their Materials of Construction, Plastic Materials of Construction, and , Plastic Packaging Systems for Pharmaceutical Use. United States Pharmacopoeia. (2017). USP Assessment of Extractables Associated with Pharmaceutical Packaging/Delivery Systems. European Medicines Agency. (2005). Guideline on Plastic Immediate Packaging Materials. EMEA CPMP/ QWP/4359/03. European Medicines Agency. (2006). Guideline on the Pharmaceutical Quality of Inhalation and Nasal Products, in Collaboration with Health Canada. EMEA CPMP/QWP/2845/00. Food and Drug Administration. (1999). Guidance for Industry: Container Closure Systems for Packing Human Drugs And Biologics, Rockville, MD. Food and Drug Administration. (1998). Guidance for Industry: Metered Dose Inhaler (MDI) and Dry Powder Inhaler (DPI) drug products.

10 Product Quality Research Institute. (2006). Safety

11

12

13

14

15

16

17

Thresholds and Best Practices for Extractables & Leachables in Orally Inhaled and Nasal Drug Products. Paskiet, D., Jenke, D., Ball, D. et al. (2013). The product quality research institute (PQRI) leachables and extractables working group initiatives for parenteral and ophthalmic drug product (PODP). PDA J. Sci. Tech. 67: 340–447. United States Pharmacopoeia. (2017). Draft of USP Polymeric Components and Systems Used in the Manufacturing of Drug Products. United States Pharmacopoeia. (2017). Draft USP Plastic Components and Systems Used to Manufacture Pharmaceutical Drug Products. Ishii‐Watabe, A., Hirose, A., Katori, N. et al. (2015). Approaches to quality risk management when using single‐use systems in the manufacture of biologics. PharmSciTech. 16 (5): 993–1001. Merseburger, T., Pahl, I., Müller, D., and Tanner, M. (2014). A risk analysis for production processes with disposable bioreactors. Adv. Biochem. Eng. Biotechnol. 138: 273–288. Parenteral Drug Association. (2014). Application of Single‐Use Systems in Pharmaceutical Manufacturing. Technical Report No. 66. ISBN 978‐0‐939459‐69‐8. United States Pharmacopoeia. (2017). USP In‐Vitro Biological Reactivity – Cytotoxicity Testing, USP In‐Vivo Biological Reactivity – Class I–VI Testing.

217

218

17  Considerations on Performing Quality Risk Analysis for Production Processes with Single‐Use Systems

18 International Standard. (2009). ISO 10993 Part 5:

19

20 21 22 23 24

25

26 27 28

29

30 31

32

33

34

35

Biological Evaluation of Medical Devices – Tests for In Vitro Cytotoxicity. European Medicines Agency. (2011). Note for Guidance on Minimising the Risk of Transmitting Animal Spongiform Encephalopathy Agents Via Human and Veterinary Medicinal Products. EMA/410/01 rev.3. European Pharmacopoeia (2016). 9e (1), 613–625. United States Pharmacopoeia. (2017). USP Particulate Matter in Injections. United States Pharmacopoeia. (2017). USP Bacterial Endotoxins Test. U.S. Government Publishing Office. (2018). Equipment Construction. Code of Federal Regulations 21§211.65 (a). European Commission. (2014). EU Guidelines forGood Manufacturing Practice for Medicinal Products for Human and Veterinary Use. Eudralex Vol. 4, Part I, Chapter 3. Brussels. ASTM International (2010). Standard test method for oxygen gas transmission rate through plastic film and sheeting using a coulometric sensor. ASTM D3985–05e1. United States Pharmacopoeia. (2017). USP Sterility Test. European Pharmacopoeia (2016). 9e (1), 185–188. ASTM International (2007). Standard test method for leakage testing of empty rigid containers by vacuum method. ASTM D4991–D4907. Antonsen, H.R., Awafo, V., Bender, J.L. et al. (2008). Sterilizing filtration of liquids. Technical report No.26 (revised 2008). PDA J. Pharm. Sci. Technol 62 (5 Suppl TR26): 2–60. International Conference on Harmonisation. (2005). ICH Q9: Guideline for quality risk assessment. Li, K., Rogers, G., Nashed‐Samuel, Y. et al. (2015). Creating a holistic Extractables and Leachables (E&L) program for biotechnology products. PDA J. Pharm. Sci. Technol. 69 (5): 590–619. Bestwick, D. and Colton, R. (2009). Extractables and leachables from single‐use disposables. BioProcess Int. 7 (Suppl. 1): 88–94. Bio‐Process Systems Alliance (2010). Recommendations for Testing and Evaluation of Extractables from Single‐Use Process Equipment. Washington, DC: Bio‐Process Systems Alliance. United States Pharmacopoeia. (2017). USP Assessment of Drug Product Leachables Associated with Pharmaceutical Packaging/Delivery systems, USP Assessment of drug product leachables associated with pharmaceutical‐packaging/delivery systems: Orally inhaled and nasal drug products. Weibing, D., Madsen, G., Mahajan, E. et al. (2014). Standardized extractables testing protocol for single‐ use systems in biomanufacturing. Pharm. Eng. 34 (6): 1–11.

36 Ball, D.J., Norwood, D.L., Stults, C.L.M., and Nagao,

37

38

39

40

41

42

43

44

45

46

47 48

49 50

L.M. (2012). Leachables and Extractables Handbook; Safety Evaluation, Qualification, and Best Practices Applied to Inhalation Drug Products. Hoboken, NJ: Wiley. Hauk, A., Jurkiewicz, E., Pahl, I. et al. (2018). Filtration membranes – scavengers for Leachables. Eur. J. Pharm. Sci 120: 191–198. Paudel, K. (2018). Investigation on fate of leachables in downstream purification of biopharmaceutical drugs. Master thesis. Hochschule Offenburg, Germany. Hauk, A., Pahl, I., Menzel, R., et al. (2017). On the fate of leachables: an introduction of a concept to investigate leachables with a “Holistic” or system approach. ECI Conference, Tomar, Portugal (8–10 May). Hauk, A. (2017). On the “Fate of leachables” in biopharmaceutical up‐stream and down‐stream processes. Vonlanthen E&L‐Summit, Berlin, Germany (19 October). Magarian, N., Lee, K., Nagpal, K. et al. (2016). Clearance of extractables and leachables from single use technology via ultrafiltration/diafiltration. Biotechnol. Prog. 32 (3): 718–724. Jahn, M. and Stebler, J. (2016). The fate of leachables during biotechnology downstream processing. Rapra E&L Europe Conference, Dublin, Irland (9–11 November). Stebler, S. (2016). Investigation on the fate of leachables during biotechnology drug substance downstream processing using diverse model compounds. Master thesis. Fachhochschule Nordwestschweiz, Switzerland. Piringer, O.G. and Baner, A.L. (2008). Plastic Packaging: Interaction with Food and Pharmaceuticals. Weinheim, New York: Wiley‐VCH. Crompton, T.R. (2007). Additive Migration from Plastics into Food: A Guide for Analytical Chemists. Shawbury, Shrewsbury, Shropshire, UK: Smithers Rapra Technology Limited. Pocas, M.F., Oliveira, J.C., Oliveira, F.A.R., and Hogg, T. (2008). A critical survey of predictive mathematical models for migration from packaging. Crit. Rev. Food Sci. Nut. 48 (10): 913–928. Crank, J. (1975). Mathematics of Diffusion. London: Oxford Science Publication, Oxford Univertsity Press. Jenke, D. (2011). A general assessment of the physiochemical factors that influence Leachables accumulation in pharmaceutical drug products and related solution. PDA J. Pharm. Sci. Technol. 65: 166–176. Atkins, P. and de Paula, J. (2006). Physical Chemistry. New York: W. H. Freeman and Company. Jenke, D.R. and Rabinow, B. (2017). Proper accounting for surface area to solution volume ratios in exaggerated extractions. PDA J. Pharm. Sci. Tech. 71 (3): 225–233.

219

18 How to Assure Robustness, Sterility, and Performance of Single‐Use Systems A Quality Approach from the Manufacturer’s Perspective Simone Biel1 and Sara Bell2 1

Merck Chemicals GmbH, an affiliate of Merck KGaA, Darmstadt, Germany MilliporeSigma, Bedford, MA, USA

2

18.1 ­Introduction

18.2 ­Component Qualification

The biggest change when moving from manufacturing with traditional stainless‐steel equipment to single‐use (SU) systems is the shift in ownership of critical quality‐ control strategies from the end user to the SU supplier (Figure  18.1). This includes things such as cleanliness, integrity, sterility, and shelf life. It is the responsibility of the end user to clearly define the user requirements and specifications that need to be met by the SU supplier. It is also recommended that the end‐user assess suppliers from both a technical and quality perspective. The stronger the partnership between the end user and the supplier, the more successful the implementation will be. Due to the high growth rates of SU technologies over the last 10 years, there are more and more SU suppliers entering the market every day. There is a lack of regulatory guidance specific to SU technologies, and for this reason the quality standards between suppliers varies greatly. This makes it challenging for end users to compare one supplier to another, and ultimately make their selection. There are several industry organizations (see also Chapter 12) working to create guideline documents, best practices, and standards to drive consistency across suppliers and make it easier for end users to implement high‐quality SU technologies that pose no impact to drug product quality, and ensure patient safety. This chapter summarizes how a supplier of SU devices assures robustness, sterility, and performance of its products. The quality approach includes component qualification, sterility validation, integrity assurance (compare also Chapter 2), stability studies, and main aspects of manufacturing and control, particulate risk mitigation, and change management.

New component identification is primarily driven by introduction of a new technology to the marketplace, product change, or end‐user need. Depending on the criticality of the component, SU suppliers may need to audit the component supplier, or even the component sub‐suppliers. As part of the audit process, the component‐manufacturing process would be reviewed to ensure there are controls in place to manage particulates, bioburden load, physical properties and specifications of the component itself, and lot to lot variability. The quality management system, nonconformance, corrective action, and preventative action (NC CAPA) and change‐ control procedures would also be reviewed as part of the audit. In addition to reviewing the quality aspects of the individual product and manufacturing process, SU suppliers should understand the component supply ­ chain. How are the individual materials of construction sourced? Are there any single points of failure in the supply chain? Do they have an effective demand planning and forecasting process, have they assessed risk related to natural disasters and are those incorporated into their business continuity plan? All components must be qualified for use in pharmaceutical manufacturing processes. The SU supplier should either perform testing or verify that the following minimum testing has been performed, prior to adding the new component to the library: ●● ●● ●●

●●

Gamma compatibility (based on maximum dosage) Statement of animal origin Absence of specific chemicals (bisphenol A, latex, phthalates, etc.) United States Pharmacopeia (USP) Class VI [2]

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

220

18  How to Assure Robustness, Sterility, and Performance of Single‐Use Systems

Manufacture

• 100% QC release test • Packaging and shipping validation ISTA 2A drop and vibration tests; package integrity, sterility, and/or functionality tests • Sterilization • Cleanliness and control of manufacturing area

●● ●● ●●

Figure 18.1  Risk‐based approach to assembly manufacture and use involves SU system design, manufacturing, and end‐user application. ISTA 2A International Safe Transit Association Standards [1]. QC, quality control. Source: Reproduced with permission of Merck KGaA, Darmstadt, Germany and/or its affiliates.

Use

• Design verification through application studies • Operator training for unpackaging/ handling/installation • General product information and assembly specific documentation • Pre-use leak testing

USP Endotoxin [3] USP Particulates [4] USP Physiochemical [5]

In addition to testing, or verification, to confirm that the new component can be used for pharmaceutical manufacturing, suppliers should also perform testing, or verification, to ensure the component is suitable for SU manufacturing. At a minimum this should include: shelf life (post gamma), bioburden (pre gamma), and bacteriostasis as well as fungistasis (post gamma). The purpose of each test is described in greater detail in Section 18.3. In 2014, BioPhorum (formerly BioPhorum Operations Group) – an industry trade group that represents leading biopharmaceutical manufacturers and SU system end users  –  published a standardized extractables testing protocol, which provided SU suppliers with a common format for conducting extractables studies and reporting results [6]. With the introduction of the BioPhorum’s standardized extractables testing protocol, SU suppliers seek to obtain these test data from the component supplier, so this also becomes a factor when selecting a component supplier (see also Chapters 2 and 8).

18.3 ­Validation of Product Design Most SU systems are configured or customized designs, meaning that they represent a unique set of parts, designed to fit into an end user’s specific process. This results in a more complex validation approach from the supplier side, for the final SU assembly, compared to a standard product. SU suppliers produce thousands of designs, so it is not feasible to validate each design to the

end user’s unique operating conditions, rather a “family approach” is taken. Representative combinations of components are selected, built into an assembly and tested, to ensure they can meet predefined acceptance criteria, bracketed, based on the fact that they undergo the same manufacturing, sterilization, and transportation processes and conditions. While SU suppliers validate the most common connections to ensure that the assembly will not leak under normal operating conditions, it is up to the end users to perform engineering testing and performance qualification using their specific operating parameters, prior to implementing the assembly for process use (Figure 18.2). 18.3.1  Sterility Validation and Quarterly Dose Audit Approach 18.3.1.1  Initial Validation of Gamma Sterilization

Validation procedures for the sterilization of SU systems via gamma irradiation are well established and based on widely used industry standards which are recognized by regulatory agencies globally: ANSI/AAMI/ISO 11137 [7], Sterilization of Health Care Products – Radiation (i) comprises: ●●

●● ●●

Part 1: Requirements for development, validation, and routine control of a sterilization process for medical devices Part 2: Establishing the sterilization dose Part 3: Guidance on dosimetric aspects, the measurement of the radiation dose.

Part 2 of ANSI/AAMI/ISO 11137 describes a method for establishing a sterilizing and verification dose with a sterility assurance level (SAL)  Frcr when the flow transition occurs. The power law of Eq. (22.7) was correlated through the all datasets of Figure 22.7 and can be used to obtain a first estimate of the mixing number. Ntm

100.7

Fr Frcr

1.24

25 (22.7)

0.8

22.3.4  Suspension Dynamics 0.75

0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22 Fr (–)

Figure 22.6  Variation of the nondimensional interfacial area, Ia, with increasing Fr for two different orbital diameters: (a) do/di = 0.35 (do = 3.5 cm, di = 10 cm); (b) do/di = 0.5 (do = 5 cm, di = 10 cm).

From this figure it is evident that the assumption is  more realistic at low Fr, before the flow transition described above takes place, but even in the highest end of Fr investigated, the difference between the measured and analytical approximation is within 4%. It is also worth noting that the interfacial area is not dependent on the amount of fluid present in the reactor, h, but only  on the value of Fr, which directly controls the ­inclination of the free surface. A detailed analysis of the natural oscillation modes of the free surface, and how these are excited depending on the operating conditions, is provided in [31], while a three‐dimensional visualization of the interfacial area is given in [32].

The critical Froude number, Frcr, can also be used to determine the shaker rotational speed associated with full suspension and homogenization of microcarriers/ cells in a shaken reactor. The ratio of the suspended to critical Froude number is plotted against the operational parameter, h/di / do /di in Figure 22.8 [34]. The inset schematics shows the flow condition present in the bioreactor depending on the combinations of fluid height, orbital, and cylinder diameters used: (i) when the fluid height is below the critical threshold, h/di / do /di 1, the toroidal vortex that controls the flow can extend to the bottom of the reactor and lift the microcarriers; (ii) when h/di / do /di 1s , the toroidal vortex does not reach the bottom of the reactor for any speed, and microcarrier suspension occurs when the swirling flow occurring after transition is intense enough at the reactor base. This explains why in the suspension Froude is Frs = 1.1 Frcr for fluid height lower than the critical one, while it tends to drift further away from the dashed reference line (Frs  =  1.1 Frcr) as the ratio h/di / do /di increases. It is worth noting that the scaling of the suspension with the critical speed can be extended to different shaken reactor geometries, including Erlenmeyer flasks ([35]).

265

22  Engineering Parameters in Single‐Use Bioreactors

Figure 22.7  Scaling of the mixing numbers with the Froude number ratio, Fr/Frcr, obtained from the measurements of Rodriguez et al. [33] (gray shaded area) and Tissot et al. [21] (data points) for very different reactor sizes and fill volumes.

400 di = 10 cm hf /di = 1.27 do = 5 cm

350

di = 13 cm hf /di = 0.58 do = 5 cm di = 16 cm hf /di = 0.31 do = 5 cm

300

di = 28.7 cm hf /di = 0.7 do = 2.5 cm di = 28.7 cm hf /di = 0.7 do = 5 cm

NTm (–)

250

di = 28.7 cm hf /di = 0.97 do = 5 cm di = 28.7 cm hf /di = 0.7 do = 5 cm di = 28.7 cm hf /di = 0.42 do = 5 cm

200 150 100 50 0 0.5

1

1.5

2

2.5

3

3.5

4

4.5

Fr/FrCr (–)

Figure 22.8  Variation of the suspended to critical Froude number ratio, Frs /Frcr, with critical height ratio, h/di / do /di , for different operating conditions [34]. Source: Reproduced with permission of Elsevier.

2 do = 1.5 cm, h = 5 cm, c = 2.5 g/l do = 1.5 cm, h = 5 cm, c = 7.5 g/l do = 1.5 cm, h = 5 cm, c = 12.5 g/l do = 2.5 cm, h = 5 cm, c = 2.5 g/l do = 2.5 cm, h = 5 cm, c = 7.5 g/l do = 2.5 cm, h = 5 cm, c = 12.5 g/l do = 5 cm, h = 5 cm, c = 7.5 g/l do = 5 cm, h = 5 cm, c = 12.5 g/l do = 2 cm, h = 3 cm, c = 2.5 g/l do = 3 cm, h = 3 cm, c = 2.5 g/l do = 4 cm, h = 3 cm, c = 2.5 g/l

1.8

1.6

1.4

1.2 Frs/Frcr

266

1

0.8

0.6

0.4

0.2

0

0

0.2

0.4

0.6

0.8

1 h di

/

1.2 do di

1.4

1.6

1.8

2

22.4  Rocking Bag

22.4 ­Rocking Bag Flow dynamics measurements obtained using laser‐ based techniques have been challenging due to the need to achieve optical access within the bag bioreactor and the requirements for bespoke mimic development. Velocity measurements, obtained by Oncul et al. [36] at single points in the flow with a hot‐film anemometer, were used to validate computational fluid dynamics (CFD) simulations at 2 and 20 l scales. They described laminar conditions at rocking rate N  = 15 rpm and reported maximum velocities of 0.2 m/s in a 2 l bag and 0.6 m/s in a 20 l bag over the course of one rock. Two dimensionless numbers can be defined for the rocking system, Reynolds and Froude, with the characteristic velocity defined as the velocity at the extremity of the bag, αNL/2, while the half‐bag length, L/2, used as the characteristic length. Reynolds and Froude numbers can therefore be defined in Eqs. (22.8) and (22.9) [37]: Re Fr

NL2 (22.8) 4 2

N2L (22.9) 2g

Marsh et al.’s [37] is one of the first studies where the flow dynamics was experimentally studied using PIV at

different rocking speeds, rocking angles, and liquid fill volumes. Figure 22.9 shows contour plots of the velocity magnitude, with superimposed vector field. The fluid moves in a parallel direction to the platform base, with the fluid bulk moving from the right end of the bag at x/L = 1 toward the center at an average velocity magnitude of 0.12 m/s. A significant effect of the rocking speed was demonstrated in this work, both on the fluid position within the bag and on the flow field. At N = 25 rpm, the fluid direction is mainly along the platform base, with axial velocity components present only at the far right end of the bag. As the rocking speed is increased, the fluid appears to occupy different portions of volume within the bag. The results obtained could be due to the fluid moving out of phase with respect to the platform, a phenomenon already described for shaken flows. The difference in fluid‐flow pattern observed at different rocking speeds generates differences in turbulent kinetic energy values measured in the bulk of the fluid. At N = 42 rpm, values of k up to 0.1 m2/s2 were measured, an order of magnitude higher than those observed at lower rocking speeds. It is also evident that a flow transition might occur between N = 33.5 and 42 rpm, demonstrated by the significant difference in turbulence levels and in the surface shape between the two conditions. It is important to study and quantify these phenomena to suitably define operating conditions depending on the process requirements.

(a)

(b)

Linear fit of liquid free surface Liquid free surface z x Calculated liquid center of mass

Figure 22.9  Flow visualization in the structural mimic. (a) Image of the fluid flow position in the mimic at a fixed angle, determination of the outer bound of the liquid and estimation of the liquid surface inclination. (b) Schematic diagram of a representative bag and fluid position [37]. Source: Reproduced with permission of John Wiley & Sons.

267

268

22  Engineering Parameters in Single‐Use Bioreactors

22.5 ­Summary and Conclusions A variety of SUB configurations are currently available for process development studies. Rigorous engineering studies are often challenging in turbulent three‐phase flows; however, these are crucial to quantify velocity, mixing, and suspension characteristics and combine such data to develop scaling correlations, and define and

validate the use of suitable dimensionless numbers. Recent studies were summarized in this chapter, where in‐depth engineering analysis of flows have led to a better understanding of transition processes, accurate quantification of local and average shear stresses, and requirements for minimum suspension speeds to inform the selection of operating conditions and facilitate translation from small‐scale reactors.

Nomenclature CFD DISMT PIV SUB do di Fr

Computational fluid dynamics Dual Indicator System for Mixing Time Particle image velocimetry Single‐use bioreactor Orbital diameter, m Reactor internal diameter, m Froude number, −

Frcr h Ia N Re Vf ν

Critical Froude number, − Fluid height at rest, m Free‐surface interfacial area, m2 Rotational speed, RPM Reynolds number, − Fluid fill volume, m3 Fluid kinematic viscosity, m2/s

­References 1 Werner, A. (2013). Large‐scale manufacturing of

2

3

4

5

6

7

biopharmaceuticals – speed up the road to market by scale up: the 6 × 15 000 l BI bioreactors. In: Biopharmaceuticals (ed. J. Knäblein and A. Werner), 527–537. Weinheim, Germany: Wiley‐VCH Verlag GmbH & Co. KGaA. Eibl, R., Werner, S., and Eibl, D. (2009). Bag bioreactor based on wave‐induced motion: characteristics and applications. Adv. Biochem. Eng. Biotechnol. 115: 55–87. Terrier, B., Courtois, D., Hénault, N. et al. (2007). Two new disposable bioreactors for plant cell culture: the wave and undertow bioreactor and the slug bubble bioreactor. Biotechnol Bioeng. 96 (5): 914–923. Baldi, S., Hann, D., and Yianneskis, M. (2002). On the measurement of turbulence energy dissipation in stirred vessels with PIV techniques. Proceedings of the 11th International Symposium on Applications of Laser Techniques to Fluids Mechanics, Lisbon, Portugal (11–14 July). Ducci, A. and Yianneskis, M. (2007). Vortex tracking and mixing enhancement in stirred processes. AIChE J. 53 (2): 305–315. Gabriele, A., Nienow, A.W., and Simmons, M.J.H. (2009). Use of angle resolved PIV to estimate local specific energy dissipation rates for up‐ and down‐ pumping pitched blade agitators in a stirred tank. Chem Eng. Sci. 64 (1): 126–143. Löffelholz, C., Husemann, U., Greller, G. et al. (2013). Bioengineering parameters for single‐use bioreactors: overview and evaluation of suitable methods. Chem. Ing. Tech. 85 (1–2): 40–56.

8 Nienow, A.W., Rielly, C.D., Brosnan, K. et al. (2013).

9

10

11

12

13

14

15

The physical characterisation of a microscale parallel bioreactor platform with an industrial CHO cell line expressing an IgG4. Biochem. Eng. J. 76: 25–36. Odeleye, A.O.O., Marsh, D.T.J., Osborne, M.D. et al. (2014). On the fluid dynamics of a laboratory scale single‐use stirred bioreactor. Chem. Eng. Sci. 111: 299–312. Abu‐Reesh, I. and Kargi, F. (1991). Biological responses of hybridoma cells to hydrodynamic shear in an agitated bioreactor. Enzyme Microb. Technol. 13 (11): 913–919. Cherry, R.S. (1993). Animal cells in turbulent fluids: details of the physical stimulus and the biological response. Biotechnol. Adv. 11 (2): 279–299. Kretzmer, G. and Schügerl, K. (1991). Response of mammalian cells to shear stress. Appl. Microbiol. Biotechnol. 34 (5): 613–616. Oh, S.K.W., Nienow, A.W., Al‐Rubeai, M., and Emery, A.N. (1989). The effects of agitation intensity with and without continuous sparging on the growth and antibody production of hybridoma cells. J. Biotechnol. 12 (1): 45–61. Ma, N., Koelling, K.W., and Chalmers, J.J. (2002). Fabrication and use of a transient contractional flow device to quantify the sensitivity of mammalian and insect cells to hydrodynamic forces. Biotechnol. Bioeng. 80: 428–437. Godoy‐Silva, R., Chalmers, J.J., Casnocha, S.A. et al. (2009). Physiological responses of CHO cells to

­  References

16

17

18 19

20

21

22

23

24

25

26

repetitive hydrodynamic stress. Biotechnol. Bioeng. 103 (6): 1103–1117. Rodriguez, G. (2016). An engineering characterisation of shaken bioreactors: flow, mixing and suspension dynamics. PhD thesis. University College London. Zhang, H., Lamping, S.R., Pickering, S.C.R. et al. (2008). Engineering characterisation of a single well from 24‐well and 96‐well microtitre plates. Biochem. Eng J. 40 (1): 138–149. Klöckner, W. and Büchs, J. (2012). Advances in shaking technologies. Trends Biotechnol. 30 (6): 307–314. Raven, N., Rasche, S., Kuehn, C. et al. (2015). Scaled‐up manufacturing of recombinant antibodies produced by plant cells in a 200‐L orbitally‐shaken disposable bioreactor. Biotechnol. Bioeng. 112 (2): 308–321. Stettler, M., Zhang, X., Hacker, D.L. et al. (2007). Novel orbital shake bioreactors for transient production of CHO derived IgGs. Biotechnol. Prog. 23 (6): 1340–1346. Tissot, S., Farhat, M., Hacker, D.L., and Wurm, F.M. (2010). Determination of a scale‐up factor from mixing time studies in orbitally shaken bioreactors. Biochem. Eng. J. 52 (2–3): 181–186. Büchs, J., Maier, U., Milbradt, C., and Zoels, B. (2000). Power consumption in shaking flasks on rotatory shaking machines: I. Power consumption measurement in unbaffled flasks at low liquid viscosity. Biotechnol. Bioeng. 68 (6): 589–593. Büchs, J., Maier, U., Milbradt, C., and Zoels, B. (2000). Power consumption in shaking flasks on rotary shaking machines: II. Nondimensional description of specific power consumption and flow regimes in unbaffled flasks at elevated liquid viscosity. Biotechnol. Bioeng. 68 (6): 594–601. Klöckner, W., Tissot, S.S., Wurm, F., and Büchs, J. (2012). Power input correlation to characterize the hydrodynamics of cylindrical orbitally shaken bioreactors. Biochem. Eng. J. 65: 63–69. Weheliye, W., Yianneskis, M., and Ducci, A. (2013). On the fluid dynamics of shaken bioreactors – flow characterization and transition. AIChE J. 59: 334–344. Weheliye, W.H., Cagney, N., Rodriguez, G. et al. (2018). Mode decomposition and Lagrangian structures of the

27

28

29

30

31

32

33

34

35

36

37

flow dynamics in orbitally shaken bioreactors. Phys. Fluids 30 (3): https://doi.org/10.1063/1.5016305. Ducci, A. and Weheliye, W.H. (2014). Orbitally shaken bioreactors‐viscosity effects on flow characteristics. AIChE J. 60 (11): http://doi.org/10.1002/aic.14608. Mancilla, E., Palacios‐Morales, C.A., Córdova‐Aguilar, M.S. et al. (2015). A hydrodynamic description of the flow behavior in shaken flasks. Biochem. Eng. J. 99: 61–66. Palacios‐Morales, C., Aguayo‐Vallejo, J.P., Trujillo‐ Roldán, M.A. et al. (2016). The flow inside shaking flasks and its implication for mycelial cultures. Chem. Eng. Sci. 152: 163–171. Weheliye, WH. (2013). Mixing, velocity and turbulence characteristics of shaken bioreactor. PhD thesis. University College London. Reclari, M., Dreyer, M., Tissot, S. et al. (2014). Surface wave dynamics in orbital shaken cylindrical containers. Phys. Fluids 26 (052104): http://doi. org/10.1063/1.4874612. Discacciati, M., Hacker, D., Quarteroni, A. et al. (2012). Numerical simulation of orbitally shaken viscous fluids with free surface. Int. J. Numer. Methods Fluids 71 (3): 294–315. Rodriguez, G., Anderlei, T., Micheletti, M. et al. (2014). On the measurement and scaling of mixing time in orbitally shaken bioreactors. Biochem. Eng. J. 82: 10–21. Pieralisi, I., Rodriguez, G., Micheletti, M. et al. (2016). Microcarriers suspension and flow dynamics in orbitally shaken bioreactors. Chem. Eng. Res. Des. 108: 198–209. Olmos, E., Loubiere, K., Martin, C. et al. (2015). Critical agitation for microcarrier suspension in orbital shaken bioreactors: experimental study and dimensional analysis. Chem. Eng. Sci. 122: 545–554. Öncül, A.A., Kalmbach, A., Genzel, Y. et al. (2010). Characterization of flow conditions in 2 L and 20 L wave bioreactors® using computational fluid dynamics. Biotechnol. Prog. 26 (1): 101–110. Marsh, D.T.J., Lye, G.J., Micheletti, M. et al. (2017). Fluid dynamic characterization of a laboratory scale rocked bag bioreactor. AIChE J. 63 (9): 4177–4187.

269

271

23 Alluvial Filtration An Effective and Economical Solution for Midstream Application (e.g. Cell and Host Cell Protein Removal) Ralph Daumke1, Vasily Medvedev2, Tiago Albano2, and Fabien Rousset3 1 

FILTROX AG, St. Gallen, Switzerland Univercells SA, Gosselies, Belgium 3  DAICEL Bioseparations – Chiral Technologies Europe SAS, Illkirch, France 2 

23.1 ­Introduction Clarification of fermentation broths (midstream) is one of the most important and disregarded steps in biotech processes. Cell culture processes have become very popular to produce therapeutics and diagnostics in the biotech world. Optimized fermentations have led to high cell densities over the recent decade. The first purification step after fermentation is the cell removal from several ml up to thousand l culture broth. The goal of the selected method is to remove the cells and cell debris as well as to reach the maximum product yield in compliance with the existing regulatory environment. Standard technologies (centrifugation, separation, membrane, and depth filtration) can no longer handle the high‐particle loads (>108 cells/ml) in an economical way. Centrifugation and separation display applications in which mechanical stress is applied to cells. They may increase releases of proteases, host cell proteins (HCPs), and other particles which have to be removed during the purification in order to exclude a negative impact on product quality and stability [1]. Membranes are very cost‐intense and it is not practical to scale them up. A depth filter with higher capacity per area would solve these issues but an increase of required footprint would be necessary. 23.1.1  Alluvial Filtration Alluvial or cake filtration, e.g. diatomaceous earth (DE) filtration, is a well‐established method in pharmaceutical industries (plasma fractionation). Until recently, however, it has not been used for single‐use cell separation, as it was unavailable as a scalable and disposable solution with all validation requirements. In 2012, the FILTRODISC™ BIO SD from FILTROX AG was the first depth filter using the advantages of alluvial technology in

a disposable format. Since then, further single‐use versions (e.g. from Sartorius Stedim Biotech and Pall) have been introduced to the market. The decision regarding the right purification system involves questions regarding process performance (product yield and quality, cell debris content, scalability, and flexibility), economics (capital investment, consumable, and maintenance costs), and existing regulatory (cleaning regulation and leachables/extractables). The singularity of alluvial, cake filtration, or dynamic depth‐filtration is the suspension of a filter aid, e.g. DE, into the feedstock before the filtration. The diatomite particles are mixed with the cells and other compressible particles. The noncompressible and high‐porous DE, and the compressible cells and cell debris perform a filter cake which avoids the blocking of the filter. This effect leads to an extraordinary solid‐holding capacity and flow through [2, 3]. The transfer from the method development stage to large‐scale processes can be easily achieved due to linear scalability. The main parameter for upscaling is the space needed for the filter cake. Lab trials can be done in a FILTRODISC™ BIO SD 2″ capsule and the volume of the filter cake calculated by the height and the area of filter cake. Multiplying by the volume of the process fermentation results in the volume of the filter cake for the production size. A scale‐up from 2″ capsule to production scale (one double 16″ module for up to 1000 l) is possible without intermediate steps. 23.1.2  Diatomaceous Earth Filtration DE or Kieselgur are skeletons of plankton deposited on lakebeds millions of years ago. It has a unique mineral morphology with a micron size of 5–50 μm, more or less pure silica and very inert. Two typical types are available: seawater and sweat water. The sweat water consists of

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

272

23  Alluvial Filtration

more than 95% one species of diatoms, has a homogenous particle size and shape, compared to seawater DE. Its main use is in the food and beverage industry, e.g. beer production. For pharmaceutical applications, special clean grades are available from two suppliers (Advanced Minerals [Celpure®] and EP Minerals [PurifiDE®]). Raw materials for current Good Manufacturing Practice (cGMP) processes need to meet standards as defined in 21CFR211.84(d)(2) and ICH Q7A. David Delvaille, biotech process sciences project coordinator, from Merck Serono said: “Diatomaceous earth is like the Swiss army knife – the all‐purpose tool for downstream processing” [4]. 23.1.3  Depth Filtration Depth filters are in the retention rate range of 50–0.04 nominal filters. They are mainly made from cellulose fibers, a porous filter aid, e.g. DE, and a positively charged polymeric resin binder. In special cases also synthetic plastic fibers or activated carbon can be integrated. Based on the composition, three filtration effects are observed: sieve filtration on the surface, depth filtration inside the filter media, and adsorption through electrokinetic effects (zeta potential) on the inner surface. 23.1.4  DAISEP MabXpure Technology Recombinant proteins such as monoclonal antibodies (mAbs) for clinical purpose contain low levels of HCPs (HCP >100 ng/mg mAb), host cell deoxyribonucleic acid (DNA 97% IgG recovery when 90% of HCP and 80% DNA remain in supernatant with DE (data not shown). If higher dilutions are applied (1 : 50 mix ratio), IgG recovery is maintained over 98% and HCP and DNA clearance start decreasing to, respectively, >64 and >78% removal. When RT is increased from 5 to 120 minutes, the HCP and DNA depletions increase, respectively, from 64 to 77% and 78 to 99% with almost complete recovery >99% (Figure 23.6). In downstream purification platforms, the traces of impurities are often removed by AEX resins or membranes under flow‐through conditions. The shorter the RT, the higher HCP depletion and IgG recovery. When all fractions are pooled, 77% HCP depletion with 95% IgG recovery is obtained for one minute RT and 74% HCP depletion with 96% recovery for IgG is obtained for two minute RT. For this CCS, 1 : 20 mix ratio is the limit for depletion. Other CCS containing ~600 000 ng/ml gives same depletion performance (data not shown). Figures 23.5 and 23.6 show that MabXpure capacity is high for contaminants. Confocal fluorescent microscopy with labeled DNA, GFP, and IgG confirmed the localization of the main families of impurities, meaning small‐size acidic homogeneously distributed within the pores, large acidic DNA adsorbed on the surface of MabXpure, and large basic IgG excluded from the porous structure. 23.3.5 Discussion Single‐use DAISEP MabXpure technology has significant flexibility for mAb contaminants removal. In filtering aid for alluvial filtration applications, MabXpure offers 1

LRV for HCP and DNA. When MabXpure is mixed with Celpure® C300, MabXpure significantly reduces turbidity (from 4000 NTU to >10 NTU), representing an ideal complement for one single clarification step. MabXpure can also be considered as enabling technology when challenging upstream conditions are occurring (high HCP and DNA concentrations). When flow‐through technology for polishing applications (co‐eluted HCP removal) is used under dynamic mode, MabXpure can replace chromatography membranes or reusable resins. Thanks to the capacity of MabXpure for contaminants and AEX‐SEC mechanisms, MabXpure permits the removal of species not depleted by other techniques. Recent studies have shown that MabXpure can eliminate one additional LRV for >60% of co‐eluted species opposed to a process using mixed‐mode resins while maintaining 5 LRV for Murine Leukemia Virus and Minute Virus of Mice [9]. This chapter focuses on single‐use process solutions. An optimized platform can reach drug substance criteria after protein A step with MabXpure and activated carbon‐based depth filters during clarification (>95% depletion for HCP and DNA with ~75% mAb yield). This platform halves the operation time and is suitable for sample preparation or when a simplified process is possible.

23.4 ­Summary and Conclusions The use of alluvial filtration (cake filtration) in midstream processing is one of the most effective, efficient, robust, and easy methods to use for cell and impurity removal. FILTRODISC™ BIO SD provides a state‐of‐ the‐art technology for this purpose. The use of filter aid over depth filter media modules is a concept, which fits perfectly with the trend for the use of disposables in bioprocessing. Diatomite filter aids suitable for cGMP

­  References

production processes are available [3]. Alluvial filtration can halve the number of filtration steps and reduce the processing time up to 75% and therefore reduces the need for highly purified rinse water [5]. It is very more up‐ and down‐scalable. Instead of a two‐step cell removal system with centrifuges, acoustic wave, others methods, and depth filters, just one step is necessary to

remove cells, cell debris, and impurities from a fermentation broth. Depth filters are able to remove impurities such as HCP, DNA, viruses, and endotoxins already, but especially in combination with MabXpure and activated carbon depth filters, the removal of HCPs reaches the next level.

Nomenclature AEC CCS CHO CT cGMP DE DNA

Anion exchange chromatography Cell culture supernatant Chinese hamster ovary Contact time Current Good Manufacturing Practice Diatomaceous earth Deoxyribonucleic acid

HCP IgG mAb NTU RT SEC

Host cell protein Immunoglobulin G Monoclonal antibody Nephelometric Turbidity Units Residence time Size exclusion chromatography

­References 1 Hassan, A.A., Mohammad, A.W., and Rahman, B.A.

(2011). Mammalian cell culture clarification: a case study using chimeric anti‐cea monoclonal antibodies. IIUM Eng. J. 12 (4): 179–187. Hurst, WE. (2005). Dynamic depth‐filtration: proof of 2 principle. http://www.advancedminerals.com/pdf/ AMC06_Bench_Scale_Proof_Princ._w._Celpure.pdf (accessed 20 October 2018). Sulpizio, T. and Taniguchi, J. (2008). Advances in 3 disposable diatomite filter aid system for cGMP Bioseparation. AFSS Annual Meeting, Valley Forge, PA (20 May). Aldridge, S. (2010). Filtration improvements yield many 4 benefits down the line. https://www.genengnews.com/ gen‐articles/filtration‐improvements‐yield‐many‐ benefits‐down‐the‐line/3487/?kwrd=contamination&p age=2 (accessed 20 October 2018).

5 ManCel Associates. (2008). Disposable Body Feed

System DBF™. Technical Bulletin.

6 Guiochon, J. and Beaver, L.A. (2011). Separation

science is the key to successful biopharmaceuticals. J. Chromatogr. A. 1218 (49): 8836–8858. Chon, J.H. and Zarbis‐Papastoitsis, G. (2011). Advances 7 in the production and downstream processing of antibodies. N. Biotechnol. 28 (5): 458–463. Smales, C.M. and Bracewell, D.G. (2014). Host cell 8 protein adsorption characteristics during protein a chromatography. Biotechnol. Bioeng. 111: 904–912. 9 Mothes, B. and Coquan, C. (2018). Case study Sanofi: effective strategies for host cell clearance in clarification & downstream operations. Biotech Week, Boston (9–12 September).

277

279

24 Single‐Use Continuous Downstream Processing for Biopharmaceutical Products Marc Bisschops, Britta Manser, and Martin Glenz Pall International Sàrl, Fribourg, Switzerland

24.1 ­Introduction The biopharmaceutical industry is witnessing a convergence of rapid changes that will affect their requirements for manufacturing strategies. This includes the increasing number of biosimilar approvals and multiple (innovator) therapies targeting the same indication (and hence the same patient population), which will result in unpredictable and often smaller annual demand. To provide more flexibility in manufacturing, the adoption of single‐use technologies for clinical and even commercial manufacturing has been very strong over the past two decades. This is partially driven by the increased process intensification (see also Chapter  14) that was achieved in – for instance – cell culture technologies. Over the course of two decades, the expression levels for monoclonal antibodies have increased by one order of magnitude. As a result, the bioreactor volume that is required for manufacturing a certain amount of product has gone down by an order of magnitude, which has enabled the use of single‐use bioreactors for manufacturing relevant amounts of product for clinical and even commercial production. A similar process intensification for downstream processing is not realistic without abandoning traditional manufacturing strategies. As a consequence, for many of the downstream processing steps there is no viable single‐use format available based on current manufacturing concepts. The most promising trend that could result in a similar increase in specific productivity for downstream processing is continuous manufacturing. The purpose of this chapter is to illustrate how continuous multicolumn chromatography is one of the key technologies to enable higher specific productivity through continuous downstream processing. Many case studies have been reported for continuous multicolumn chromatography, but the vast majority of these

involve proof of concept at bench scale only. In this chapter, we will discuss the concept of multicolumn chromatography including scale‐up to clinical and commercial manufacturing.

24.2 ­Continuous Multicolumn Chromatography Chromatography is one of the workhorses in the purification of recombinant proteins and monoclonal antibodies. A drug substance‐manufacturing process for a monoclonal antibody product includes typically three chromatography steps. However, chromatography and ion exchange are not exclusively used for the purification of proteins. Large‐ scale ion‐exchange processes have been applied to the recovery and purification of food processing, hydrometallurgical applications, and a wide variety of chemical processes as well as other industrial biotechnology products including antibiotics and amino acids. In many of these applications, continuous manufacturing has been implemented over the past decades, and the most ­successful continuous ion exchange and continuous chromatography applications use multicolumn chromatography. In this approach, the stationary phase (ion‐ exchange resin or adsorbent) is kept in multiple smaller columns and the process liquids are directed to each of these columns. The feed is applied on at least one of the columns continuously and the effluent of this primary load column(s) is passed on another (set of ) column(s) so that the primary column(s) can be overloaded beyond dynamic binding capacity. As soon as the primary load column is saturated, the feeding continues on the next column and the first (primary load) column is subjected to all subsequent (non‐load) steps such as washing, eluting, and

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

280

24  Single‐Use Continuous Downstream Processing for Biopharmaceutical Products

cleaning. Once equilibrated, the column is then added to the tail of the load step again. This concept of continuous loading in a multicolumn chromatography system is schematically shown in Figure 24.1. Various mechanical concepts for multicolumn chromatography have been implemented in the industry. Chin and Wang [1] provided a broad overview, classifying multicolumn chromatography systems in those with an integrated rotating valve system and those with a distributed valve system. Both concepts have found large‐ scale applications. Rotating valves can – for instance – be found in the lysine purification plant as described by van Walsem [2], the vitamin C purification plant described by De la Fuente et al. [3], and in all variants of the Sorbex technology as originally developed by Universal Oil Products [4]. For biopharmaceutical applications, the concept of a rotating valve presents significant challenges related to cleaning. As a result, the distributed valve concept has become the dominant design for multicolumn chromatography systems in the biopharmaceutical industry. This concept also provides more flexibility as it supports asynchronous switching and hence it allows performing more complex purifications with a limited number of columns. One of the key advantages of continuous multicolumn chromatography processes is the increased ­specific productivity. This allows producing the same or larger amounts of product with a significant reduction in adsorbent volume. It has often been stated that the volume of the relatively expensive chromatography adsorbents is the key limiting factor in establishing a viable disposable downstream processing platform. This limitation can be overcome by the use of ­continuous bioprocessing technologies.

(a)

24.3 ­Single‐Use Continuous Downstream Processing 24.3.1  Continuous Chromatography for Fed‐Batch Processes The significant increase in specific productivity that can be obtained by using multicolumn chromatography has resulted in the design, development, and commercialization of various multicolumn chromatography solutions. The Cadence® BioSMB technology that is commercialized by Pall Biotech (Port Washington, NY) is unique in the sense that it is the only technology that offers a fully single‐ use flow path. With this, the Cadence BioSMB technology offers a viable solution for single‐use chromatography to serve clinical and commercial manufacturing. The first scale‐up data on a Cadence BioSMB Process system were generated at Merck & Co. Inc. (Kenilworth, NJ) [5]. The feed material for the case study was generated in a 500 l fed‐batch bioreactor, yielding 5.6 g/l monoclonal antibody in solution. This case study illustrated the entire development workflow for a Protein A capture process using Kaneka KanCapA™ affinity adsorbent. To shorten the development timelines as much as possible, a model‐ assisted design approach was used as described earlier [6]. This approach involves the use of a few batch breakthrough curves to derive the relevant process design parameters for the multicolumn chromatography process. Based on these parameters, the typical continuous chromatography response curve was generated (Figure 24.2). This response curve indicates the required total contact time in the load step that is required to achieve a certain protein load on the chromatography columns. Above the response curve, the total contact time is sufficient to allow

(b)

Feed

Equil.

Drain

Waste

(c)

Feed

Equil.

Drain

Waste

Wash

Feed

Drain

Figure 24.1  Conceptual diagram of a multicolumn chromatography process: (a) beginning of a load step: the feed is applied on a primary load column which then feeds the secondary load column, (b) end of a load step, and (c) upon completion of the load step: the primary load column is being washed while the secondary load column has become the primary load column.

24.3  Single‐Use Continuous Downstream Processing

Required contact time in a BioSMB process 6.00

Required contact time (min)

5.00 4.00 3.00 2.00 1.00 0.00 20.00

25.00

30.00

35.00

40.00

45.00

50.00

55.00

60.00

Protein load on column (gl)

Figure 24.2  Cadence BioSMB Process response curve: above the curve, complete capture of the monoclonal antibody is expected, whereas below the curve, loss of monoclonal antibody will occur due to too short contact times. The data point at 46.4 g/l and 3:04 min contact time represents the process conditions that were selected for the process scale run.

100% Experimental capture efficiency (–)

Capture efficiency (–)

100% 95% Exp: 3.19 – 3.35 min Exp: 2.24 – 2.40 min

90%

Model: 3.32 min Model: 2.32 min

85% 80% 75% 30.0

35.0

40.0

45.0

50.0

55.0

60.0

65.0

70.0

Load capacity (g/l)

75.0

95% 90% 85% 80% 75% 75%

80%

85%

90%

95%

100%

Predicted capture efficiency (–)

Figure 24.3  Experimental data on the Cadence BioSMB PD system to challenge the model‐assisted process design approach. Left: experimental capture efficiencies versus load capacity at different load contact times; Right: parity plot between predicted and experimental capture efficiency.

99.5% of the antibody to be captured. Below the response curve, the contact time is too short and some material will be lost in the breakthrough (drain) of the load step. In order to verify the model‐assisted process design approach, small‐scale experiments were conducted on a Cadence BioSMB PD system at bench scale. The performance of these experiments corresponded accurately to the model prediction. This is shown in Figure 24.3. The final process conditions for the process scale run are also shown in Figure 24.2.

The process conditions and results for the process scale run are summarized in Table  24.1. The process involved three columns in the load step and two columns to cover all non‐load steps (wash steps, elution, sanitization, and equilibration). The continuous chromatography experiment was also performed on a Cadence BioSMB PD system at small scale, using exactly the same critical process parameters. In addition to this, a batch reference process was performed using the same feed solution and the same adsorbent.

281

282

24  Single‐Use Continuous Downstream Processing for Biopharmaceutical Products

Table 24.1  Experimental conditions and results for the case study performed on the Cadence BioSMB Process 350 system, including batch reference and scale‐down study on the Cadence BioSMB PD system. Batch reference

Cadence BioSMB PD

Cadence BioSMB process

Flow rate

4.0 ml/min

37 l/h

Column dimensions

11.2 mm ID

14 cm ID

5 cm H

5 cm H

5 × 5 ml = 25 ml

5 × 0.77 l = 3.85 l

Contact time

Column volume 4:00 min

3:04 min

3:04 min

Binding capacity

30 g/l

46.4 g/l

46.4 g/l

Yield

98%

97%

97%

Aggregate level

0.45%

0.72%

0.65%

DNA removal

n.a.

4.2 log

5.0 log

HCP removal

2.4 log

2.6 log

2.5 log

Specific productivity

16 g/l/h

56 g/l/h

56 g/l/h

During the process scale work, over 400 l of clarified cell supernatant was purified in approximately 12 hours. In total, 13 chromatography cycles were performed and process consistency was maintained throughout. The small difference in aggregate levels between the batch reference process and the continuous alternatives was most likely related to a different sample treatment. All other impurity data suggest that the impurity removal in the continuous process was very similar to the batch reference process. One of the key results of the case study was that the entire workflow from redefining a batch process until the process scale run covered a period of no more than three weeks. During this period, the process was characterized, the Cadence BioSMB Process 350 system was installed in the pilot plant at Merck & Co. Inc. (Kenilworth, NJ), and people were trained to operate the system. 24.3.2  Continuous Chromatography for Perfusion Processes The increased specific productivity of continuous multicolumn chromatography makes it a very attractive technology for medium and high titers as often observed in modern fed‐batch processes. On the other hand, however, linking a continuous capture chromatography process to a continuous perfusion bioreactor also seems to be an obvious combination to establish a fully integrated continuous biomanufacturing platform [7, 8]. The key challenge for the affinity chromatography step in combination with a perfusion bioreactor is to maintain bioburden control during longer production campaigns. Perfusion (see also Chapter  14) campaigns can

last as long as 60 days, although longer campaigns have become quite common as well. Although downstream processing is not a sterile process, bioburden control is critical in the capture process as it involves rich growth medium. As a result, the presence of bioburden in the capture step could escalate quickly. In order to address this risk, a dedicated design of the single‐use assembly for Cadence BioSMB Process 80 system has been established that allows operating the ­affinity capture process in a bioburden‐controlled environment. The single‐use assembly is manufactured as an integrated functionally closed system and is supplied gamma irradiated and double‐bagged to allow installation while maintaining zero initial bioburden. Aseptic connectors allow connecting columns, process solutions, and product/waste containers without breaching the functionally closed state. Preliminary results on the use of the Cadence BioSMB Process 80 with this dedicated single‐use assembly were presented by Ötes et  al. [7]. The work involved a 40× scale‐up from 5 ml columns to pilot scale. The specific productivity for the pilot‐scale experiments was in the range of four to five times higher than in the reference batch process, while maintaining consistent product quality attributes. The experimental work included a test in which the system was operated for the fully automated capture of monoclonal antibodies from cell supernatant during 10 consecutive days. During this long‐term test, no evidence of bioburden was found, suggesting that the technology is capable of running longer campaigns without bioburden ingress. Additionally, product quality remained very consistent and (at least) equal to the batch reference process.

­  References

24.4 ­Summary and Conclusions Continuous downstream processing results in much higher specific productivities. This allows translating a batch process into a much more compact continuous process, delivering the same product output. As a consequence, single‐use format becomes a viable option also for downstream processing. Scale‐up of continuous multicolumn chromatography to pilot‐scale and commercial‐scale production has been shown straightforward and consistent using the Cadence BioSMB technology. This is facilitated by the fact that the architecture of the process development system and the process scale system are identical, which minimizes risks associated with dead volumes and other scale‐dependent effects. In addition to this, the similari-

ties in architecture also enable portability of the phase files that contain all instructions to run the systems across various scales. For clinical manufacturing and/or manufacturing of smaller amounts of (niche) products, the added value of continuous chromatography is predominantly delivered through the increased specific productivity, which en­ables significant savings in relatively expensive affinity adsorbent. In addition to this, the single‐use design of the entire flow path allows rapid changeover and reduces risks that would otherwise impact facility utilization. For routine manufacturing, the value of continuous multicolumn chromatography is among others delivered through its increased capacity utilization. In addition to this, having a lower installed volume of affinity adsorbent also reduces the impact of operational risks.

­References 1 Chin, C.Y. and Wang, N.H.L. (2004). Simulated moving 2

3

4

5

bed designs. Sep. Purif. Rev. 33 (2): 77–155. van Walsem, H.J. and Thompson, M.C. (1997). Simulated moving bed in the production of lysine. J. Biotechnol. 59: 127–132. De la Fuente, R., Pennings, M., and Bisschops, M. (2007) Continuous multicolumn ion exchange for vitamin C production. EFB Workshop on Downstream Processing, Delft, The Netherlands (10–11 May). Broughton, D.B. and Gerhold, G.C. (1961). Continuous sorption process employing fixed bed of sorbent and moving inlets and outlets. US Patent 2, 985,589. Brower, M. (2016). Scale‐up of continuous chromatography using Cadence BioSMB process system. Bioprocess International Conference, Cambridge, MA (4 October).

6 Bisschops, M. and Brower, M. (2013). The impact

of continuous multicolumn chromatography on biomanufacturing efficiency. Pharm. Bioprocess. 1 (4): 361–372. Ötes, O., Flato, H., Vazquez Ramirez, D. et al. (2018). 7 Scale‐up of continuous multicolumn chromatography for the protein a capture step: from bench to clinical manufacturing. J. Biotechnol. 281: 168–174. Bisschops, M. (2018). Producing safe biotherapeutics 8 using continuous downstream processing technologies. Continuous Biomanufacturing: Achievements and Challenges for Commercial Implementation, Oxford, UK (22 June).

283

285

25 Single‐Use Technology for Formulation and Filling Applications Christophe Pierlot1, Alain Vanhecke1, Kevin Thompson2, Rainer Gloeckler3, and Daniel Kehl3 1

 Pall Biotech, Hoegaarden, Belgium  Pall Biotech, Portsmouth, UK 3  Swissfillon AG, Visp, Switzerland 2

25.1 ­Introduction The biopharmaceutical production of drugs typically consists of two main process sequences that often occur at distinct production sites: the drug substance (DS) manufacturing and the drug product (DP, the final drug for patient administration) manufacturing [1]. The DS manufacturer has an expertise in cell expansion and/or infection followed by a series of purification steps ­ending in a purified, conditioned, and filter‐sterilized bulk. This DS bulk is typically subdivided into smaller fractions to be sent to the receiving DP manufacturers. In some cases, the DS bulk is frozen, prior to shipment, or final formulation into DP. The DP manufacturers have a specific set of technical expertise and process capabilities enabling them to compound the active ­pharmaceutical ingredient (API) with other excipients, such as salts, buffers, stabilizers, surfactants, and where applicable, bacteriostatics, into a formulated DP. When the DP is heat sensitive, the formulated biopharma­ ceutical DP is sterilized by filtration after which it is  accurately filled into its final drug container, also called primary container, through a series of aseptic techniques. These primary containers can range from syringes, bottles, vials, ampoules, or cartridges and are aimed to protect the drug from external influences thus maintaining its desired potency and efficacy. The choice of the primary container type is driven by patient safety, drug quality, patient administration preference, and cost [2]. This chapter will specifically go through DP formu­ lation and filling challenges, talk through end‐user ­process requirements, and give an overview of single‐ use technologies that can be used to facilitate process­ ing in this area.

25.2 ­Challenges in Formulation and Filling The fragile nature of therapeutic proteins is of high con­ cern at the later stages of biopharmaceutical production where purified and concentrated DS and final product are handled. In recent years, pharmaceutical companies have turned increasingly to highly concentrated protein formulations [3]. The main driver for this being conveni­ ence of subcutaneous instead of intravenous patient administration and related cost‐benefits of self‐adminis­ tered as opposed to health‐care professional‐adminis­ tered drugs. As the subcutaneous DPs are only allowed to be administered in a specific dose range not exceeding 1.5 ml, as per Food and Drug Administration (FDA) rules, this has consequently led to increasingly high con­ centrations of new monoclonal antibody (mAb) formula­ tions greater than 100 mg/ml [4]. These DP formulations tend to be viscous and are prone to form aggregates due to shear or microcavitation as can be generated by pumps and valves that can impact drug efficiency, stability, and safety during processing [5]. The aim of DP manufacturers is to formulate, fill, and finish a DP that meets its critical quality attributes (CQAs) ensuring patient safety, drug stability, and drug potency. The robustness of such DP formulation is as important as the manufacturing robustness and is based on prior knowledge, risk assessments, quality‐ risk management, and statistics [6]. The drug manu­ facturers will formulate to a defined specification, which enables them to accurately reach a pH value, ionic strength, chemical composition, concentration, homogeneity, and formulation temperature. The type of excipients to be added will be intimately linked to

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

286

25  Single‐Use Technology for Formulation and Filling Applications

the finished DP, liquid or solid phase (e.g. lyophilized product), for which specific excipients will enable the products to be maintained or transformed, respec­ tively, to the desired physical phase. Once the final formulation is homogenous, it should be sterilized and filled without negative impact on the stability or functionality of the DP and with minimal product loss. To achieve this, the following equipment is necessary: thawing equipment, pooling or conjugation equipment (e.g. vaccines), mixing equipment possibly equipped with thermal control unit, automated formula­ tion control and dosing unit, filtration step/s, and final filling machine. The duration of processing is a major economic con­ sideration for any DP manufacturer and should therefore be minimized. However, the processing time will be dic­ tated by the DP itself due to the precise boundaries that must be complied with, such as: storage temperature, sensitivity to light, oxidation rate, mixing rate linked to shear stress, etc. Beyond the simplistic view of mixing and filtration, the trend toward complex formulations raises very spe­ cific challenges to overcome during the final formula­ tion and filling operations. Outsourcing of DP processing is therefore a common practice as it requires specific technical expertise, process equipment, and capabilities to effectively address these challenges [7]. The complex formulations can be categorized as follows: suspen­ sions (e.g. emulsions, liposomes, and nanoparticles), drug conjugates (e.g. antibody–drug conjugates, ADCs), crystallized proteins, and mAbs (e.g. high‐concentra­ tion mAbs). Some of these processes, such as ADCs (see also Chapter 27), present important toxicity or potency challenges that underline the need for specialized DP manufacturers. The high degree of outsourcing in the DP area means these facilities operate as multiproduct ones that maxi­ mize manufacturing capacity and utilization to make this an economically viable model. Traditionally, the majority of product surface contact parts in processing were made of stainless steel, glass, or elastomeric material. Due to their mechanical properties, these materials have been used in a reusable manner which implies end users need to validate cleaning procedures and sterilization methods (steam autoclave or steam­ ing in place), and schedule regular preventive mainte­ nance to mitigate cross‐contamination risks associated with equipment reuse. The evolution toward utiliza­ tion of polymeric materials for single‐use processing enabled the conception of entire closed processes with plastic surface contact material that are required to be disposed of after every batch. These single‐use systems (SUSs) are delivered as preassembled, presterilized manifolds thereby eliminating the need for cleaning,

sterilization, and maintenance by the end user. Furthermore, adoption of single‐use technology has improved manufacturing efficiency, enabled faster project execution, and allowed for greater manufactur­ ing flexibility. Single‐use technologies are today available at all steps of the process including final formulation and filling operations. Interestingly, adoption at commercial scale asked for major improvements in consistency and relia­ bility compared to rather manual operation at clinical scale. Leading SuS suppliers encouraged this shift and offer their engineering expertise in the design of indus­ trialized single‐use technology platforms, with a high degree of process automation and a shared onus on qual­ ity and quality‐risk management with the end users.

25.3 ­End‐User Requirements The starting point for the discussion of end‐user needs with SUS suppliers is a well‐defined user requirement specification. We subsequently describe the view of Swissfillon AG Visp, Switzerland, a custom manufac­ turer in the area of final fill activities using SUS from compounding over sterile filtration down to the filling needle. The most important Swissfillon user requirement was to have a fully single‐use product contact flow path that would span from formulation, filtration, and final filling. It is understood that this implies the flow path would span across a variety of physical barriers neces­ sary to maintain different processing environments to safeguard adequate product quality and operator pro­ tection. The primary objective behind usage of a fully single‐use flow path is to avoid product‐specific clean­ ing and corresponding cleaning validation as this would imply major problems in analytical detection limit eval­ uation for highly toxic substances to be filled. Indeed, such products indicate toxic exposure levels of less than or equal to 1 ng/m3 of air, which would imply substan­ tial efforts to develop, qualify, and effectively repeat cleaning procedures for each type of product or for dif­ ferent batches. The Swissfillon selection of the appropriate supply partner for such SUS systems included the following considerations as part of the decision tree: ●●

●●

●●

Experience in the development of critical single‐use manifolds including final filling needles. Supplier flexibility to accept specified components as part of such systems including preferred materials of construction, connectors, and filters. Packaging quality and consistency with proof of trans­ port validation.

25.4  Quality by Design ●●

●●

●●

●●

●●

●●

●●

●●

●●

Assemblies to be manufactured and double bagged in International Organization for Standardization (ISO) 7 environment. Sterilization validation, capacity, and backup through usage of multiple validated irradiation sites. Design concept, consistency, and ease of unpacking and installation. Certifications and documentation fulfilling compen­ dial requirements and international standards. Extractables information to complete toxicity assessments. Supplier’s manufacturing capacity, redundancy, and capabilities to meet product continuity assurance and consistency in supply chain performance. Operational performance for design creation, design changes, minimal order quantities, and delivery time. Validation services to assist in bacterial‐retention studies and validation of filter integrity testing, adsorp­ tion, compatibility, and process‐specific extractables and leachables. High assurance of SUS integrity with capability for SUS integrity testing at the supplier and/or enabled by supplier design or technologies at the end users’ side.

Overall, it should be noted that as a contract manufac­ turing organization, Swissfillon is committed to finding a flexible and responsive partner that meets the highest standards in terms of quality and operational perfor­ mance in order to fulfill its obligations to their respective customers. After selection of the preferred supply part­ ner, the initially proposed solution was further improved based on a failure‐mode effects analysis that identified risks of failures to be addressed through further design improvements. One particular output of this analysis is pointed to a high risk in the SUS unpacking and installation proce­ dure. The supplier proposed further packaging improve­ ments and the creation of supportive hardware for the final filtration stage ensuring consistent installation and ergonomy of the SUS for this critical application. This solution was then trialed in a prototype evaluation in conjunction with the final filling machine equipment to verify compliance to objectives. Overall, a competent partner was found in Pall Biotech that fulfilled the user requirements as discussed above. Their primary SUS manufacturing plant for European customers is located in a recent facility located in Medemblik, the Netherlands. This SUS manufacturing plant is using a qualified Class 7 cleanroom in operations according to ISO 14644 with dedicated well‐trained per­ sonnel under a Quality System certified to ISO 9001 (quality management) and ISO 14001 (environmental management). This manufacturing plant has been audited as part of the supplier approval process by Swissfillon.

25.4 ­Quality by Design In traditional bioprocessing, the end user has a great level of ownership over equipment design, construction, integrity, operation, and maintenance. In contrast, with end‐to‐end closed SUSs, a larger piece of ownership now resides with the SUS supplier. The supplier has to man­ age the diverse nature of SUSs, especially when these include a variety of subcomponent suppliers, and give assurance of the supply chain, quality (certification), engineering, manufacturing, quality control, steriliza­ tion, packaging, transportation, and final operational performance of the systems. As described in Section 25.3, the end user will list his user requirements specification and clarify the intended application and process for what the SUS will be used. This allows the SUS supplier to align his proposal from component selection, assembly design, manufacturing, packaging, sterilization, validation, certification, and shipping qualification in agreement with the end‐user requirements. The objective is for end users to under­ stand the SUS life cycle from design, manufacturing, and validation so it allows them to risk assess a supplier’s pro­ posal for the desired disposable product and/or system [8]. It must be noted that due to the relatively high degree of customization, this will not be a simple task. The end user has to create his risk assessment need, and then has to identify and carefully evaluate all eventual risks that a supplier proposal presents to the intended application. Finally, understanding the overall consistency in SUS supplier’s controls, such as junction‐to‐junction leak testing, is of essence to obtain repeatable product quality in routine manufacturing. Throughout such clarification process, a great amount of information is to be shared from the SUS supplier to end user and vice versa which can be reached most effec­ tively through a collaborative approach or even a close partnership to successfully reach objectives. For the most complex matters, such as, for example, extracta­ bles, particulates and integrity concerns of SUSs, indus­ try associations, and communities, such as Bio‐Process Systems Alliance and BioPhorum Operations Group, will be of considerable help in understanding how to approach and what to expect from suppliers to overcome challenges that have impeded single‐use adoption in the past [9, 10]. According to the latest bioprocessing professional’s community survey, the increased understanding of the subject and ability to reuse supplier data have led to higher satisfaction levels among the community helping them to answer regulatory scrutiny [9]. Overall, it can be concluded that thorough understanding of end‐user product, application, and process is the starting point for SUS suppliers to consistently conceptualize the complex

287

288

25  Single‐Use Technology for Formulation and Filling Applications

SUS within an engineered and validated design space that will facilitate end‐user risk assessment, minimize validation requirements, and allow for safe and qualita­ tive DP manufacturing within predefined critical process parameters ensuring the CQAs are met throughout the whole life cycle of the DP.

25.5 ­Hardware Design and Usability From a contract manufacturer point of view, Swissfillon has been focusing on designing a flexible infrastructure with latest technologies and with high turnover objectives to strengthen its core competitive advantage being to deliver drugs on time to market. This time to market fac­ tor influences the design of the final filling machine and of the necessary side‐equipment such as mixer, pumps, fil­ tration boards, and biocontainer surge control. As an example, this relates to the flexibility of the filling equip­ ment to fill multiple primary container formats. The design of equipment should inevitably work with space restrictions of the cleanroom as its size is limited as it is proportional to cost of operations and cost of the plant. Any equipment supplier has to guarantee quick change over times in combination with the possibility to avoid any intense cleaning steps. Additionally, it has to be guaranteed that isolator technology will be used in com­ bination with ready‐to‐use materials, such as stoppers or fluid path assemblies, to ensure that short preparation time for the next batch is achieved. To fulfill the high potency or high toxicity DS han­ dling, the final filling machine had to be placed in an ISO 7 class cleanroom with negative pressure conditions for environmental safety in case of a spill event, whereas the inside of the isolator is running under overpressure con­ ditions to guarantee product safety. This design clarifies why cleaning had to be reduced to a minimum as waste water would have to be handled as contaminated and therefore be incinerated in a special oven not easily accessible. The installed final filling machine is a “Multi Use Filling Machine” by Optima, guaranteeing shortest changeover from syringe to vial to cartridge filling by exchanging only a limited amount of format parts. The reduction of human interactions was achieved through usage of robot technology. Additions of ready‐ to‐use materials from the outside, such as stoppers or caps, are performed using real‐time transport protocol‐ port technology. Alternatively, such a port will also be used to incorporate the final filling assembly (Figure 25.1), including its surge biocontainer, pump tubes, and the filling needle, into the filling machine. To connect the fil­ tration and filling manifold that are, respectively, in two different cleanroom environments, a sterile connector device is being used in ISO 7 cleanroom.

Figure 25.1  Supportive hardware for the Pall Biotech final filtration single‐use assembly.

The isolator, which is part of the filling machine, must satisfy two important requirements. First, the isolator has to ensure sterility after a performed decontami­ nation step using H2O2. Second, the isolator has to guarantee personal safety if a high‐potent/high‐toxic material will be filled. As an additional point, the canals for the exhaust air of the isolator have to be equipped with a double bag in/bag out Hepa filter (H14). All exhaust air canals connected to the filling unit must be cleanable by installed spray nozzles using hot water for injection. The definition of a project scope, the execution, and the qualification require the end user and filling machine supplier to work in a logical sequence. In this case, the V model was used as an excellent basis for design control and tracking of design changes throughout the proceed­ ings of the project. Any of these proceedings are tightly monitored by the end user to ensure that work done by the supplier matches the scope and can be leveraged to

25.5  Hardware Design and Usability

facilitate implementation, verification, and operation of the equipment. Indeed, through this way installation and operation qualification should not be repeated as it would leverage site acceptance testing. Such an approach leaves the end user to de‐risk project proceedings before construction of the machine and to limit time spent on qualification on site. Indeed, the major on‐site activity for the end user is, in addition to the “site acceptance test,” the performance qualification. The two major per­ formance qualification activities are: decontamination performance using H2O2 and the primary packaging‐ dependent mediafill using Tryptic soy broth, and the necessary filtration and filling manifolds to proof sterility of the fully integrated system. Other ancillary equipment includes: compounding isolator, mixing biocontainer equipment, and filtration panel. For the compounding of high‐potent, high‐toxic API, a single‐use isolator (Figure 25.2) as supplied by ILC Dover or Lugaia has to be installed to guarantee personal safety. Important consideration in the choice of the solu­ tion was to connect or disconnect in the most contained way. It should be noted that today’s disconnection systems only guarantee an operational exposure level of 1 μg/m3 of air, so improvement is still necessary. Furthermore,

the location of the weighing scale outside of the isolator does not help in achieving precise weighing results. The next step being to add the content of the powder bag to the compounding bag in a closed manner to allow for mixing. After adequate homogenization, the com­ pounding bag has to be connected to the filtration mani­ fold using an appropriate connector. As the filtration manifold is the most critical system guaranteeing the final sterility of the product, this manifold had to be designed in a simple, safe, and reproducible manner allowing for consistent installation regardless of operator and their respective training level. As a perfect solution for Swissfillon, a shadow board (Figure 25.3) was developed together with Pall, indicat­ ing the installation points of the manifold but also the clamp identification by numeric code and color code, with icons indicating filter and biocontainer position. The shadow board allows even inexperienced/new oper­ ators to set up the filtration manifold safely in less than two minutes. The shadow board together with the surge control allows an operator to semiautomatically prime the filtra­ tion manifold after performing a leak test of the single‐use

Figure 25.2  Single‐use isolator for safe operator handling of toxic active pharmaceutical ingredients.

Figure 25.3  Swissfillon’s shadow board indicating the installation points of the manifold and clamp identification.

289

290

25  Single‐Use Technology for Formulation and Filling Applications

Figure 25.4  Allegro final filling needle into Optima final filling machine.

assembly. To prime the filtration system, liquid is pumped into the system via surge control and adjacent peristaltic pump till the pressure increases to its upper limit. The air in the systems is discarded into small biocontainers incor­ porated into the filtration manifold. Clamps are closed again and after a second fill step the upper filter is totally filled with product solution. In a next step, the same pro­ cedure allows to fill the second filter too. As a last step, the filter integrity for both installed sterile filters will be veri­ fied using the Palltronic® Flowstar LGR leak and integrity test instrument. The last step before the filling process can start is the final connection of the filtration and filling manifold through a single‐use sterile connector followed by the priming of the filling manifold down to the single‐ use needle (Figure 25.4). The combination of an Optima MultiUse filling system and Pall Biotech as the supplier of the additional equip­ ment and the SuSs allowed Swissfillon to run nine suc­ cessful media runs (sterility runs) without any failure or lost batch.

25.6 ­Single‐Use Technology, Arrangement, and Operation The multiproduct nature of DP manufacturing, trend for smaller drug batches, shorter time to market pres­ sure, and new drug therapies reimbursement uncertain­ ties ask for flexible and cost‐effective manufacturing. Suppliers have addressed this cost and flexibility demand

by offering SUSs for formulation, filtration, and filling steps that can easily be connected or disconnected though usage of single‐use aseptic devices. The presteri­ lized SUSs are often referred to as ready‐to‐use as it only requires unpacking and installation for use, although some elements might require a preconditioning step such as a flush. Mixing is an important unit operation in DP process­ ing and the selection of a candidate mixing technology is typically based on a combination of process performance and ability to preserve drug quality features. A broad range of technologies is available to achieve homogeni­ zation of a solution (see also Chapter 3). Mostly, the sin­ gle‐use biocontainer design includes a stirrer although in rare occasions one could also achieve homogenization through recirculation with a pump or oscillation of the entire biocontainer. The ubiquitously used stirrer mixing design can be segmented into two main categories: bot­ tom‐mounted mixers and top‐mounted mixers. These can further be differentiated in stirrer type and coupling principle which conveys energy from an external source to the stirrer. Depending on the drug type, users will have to balance the process performance vs drug quality‐retention aspects. As an example, if having mAbs or related frag­ ments antigen binding, shear and cavitation stress is of particularly higher concern compared to non‐antibody‐ related proteins or small‐molecule DPs. Shear and cavi­ tation can lead to aggregates which result in filter fouling that can entail severe process delays or failures. Experimental studies with mAbs have confirmed that these molecules are particularly prone to aggregation when exposed to moving parts, particularly in the form of material‐to‐material contact [11]. On top of that, avoiding material‐to‐material friction is of importance to minimize additional particulate creation during mix­ ing. Therefore, it should not be surprising that the fric­ tionless levitating impeller mixing technology is the mixer of choice from late‐stage DSP steps to final DP formulation. Some formulations will prescribe nitrogen sparging followed by a nitrogen overlay prior to start the formulation to eliminate protein–air interactions that could in turn lead to protein denaturation and aggrega­ tion. It should be noted that final filtration offers a final opportunity to remove such aggregates. Sterilization of biopharmaceutical drugs is commonly achieved through usage of sterilizing‐grade filtration, as heat sterilization would degrade and inactivate pro­ tein structure. The final sterilizing‐grade filtration in single‐use process arrangements should preferably be in line as close as possible to the final filling point and, wherever possible, downstream of aseptic connectors [12]. Traditionally, bulk filtration has been used into mobile bulk vessels that would be transported to the

25.6  Single‐Use Technology, Arrangement, and Operation

filling station, but this entails a high degree of aseptic connections that can add unnecessary sterility risks. The selection of a final sterilizing‐grade filter should be based on: (i) microorganism‐retention capacity and generation of a sterile effluent, (ii) compatibility with the DP and process conditions, (iii) tendency not to adsorb formulation components and must not add extractables to the process stream that would impart the patient safety and/or drug quality profile, and (iv) sterilization compatibility [13]. To ensure the final filter would comply in rendering a sterile effluent, drug manufacturers are asked to confirm the integrity of sterilizing‐grade filters (see also Chapter 10) by subjecting a pre‐wetted filter to a bubble point, forward flow, water intrusion, or pressure hold test. An integrity test is always associated with a high pressure being applied to the filter system on the USP side with pressures of 1.5–3.9 bar (for sterile filters with a pore size of 0.2 μm) depending on the test procedure [14] and the wetting fluid. The filter integrity test objectives are to confirm: integrity of filter media (no holes, defects, etc.), verify correct filter installation (no missing O‐ring, etc.), and confirm no filter damage occurred during fil­ tration (post use test). When switching to fully dispos­ able technology, users should carefully risk assess if traditional practices can safely be transferred to a poly­ meric, multi‐junction disposable setup. To illustrate this, let us look at a case study where one was traditionally used to do bubble point filter testing followed by filter drying with dry heated air. The high bubble point test pressure and downstream pressurization of the flow path to the same extent requires significant design alterations to absorb such pressure relief. The filter drying practice requires higher than bubble point pressure and pre­ heated air which ask for another significant design con­ sideration to manage continuous air flow throughout the system, as well as the air quality. It goes without saying that any of these two practices expose the closed poly­ meric assembly to high stress which could be avoided if using forward flow testing and other wetting strategies, respectively. Post‐use integrity testing on sterilizing‐grade filters is a broadly accepted regulatory obligation; on the other hand, industry debate exists on the necessity for a pre‐ use integrity test after sterilization. Indeed, the European Union Annex 1 guidelines set this pre‐use post sterilization integrity test (in short, PUPSIT) requirement out in 2008 and the latest draft revision will very likely retain this point [15]. The main end‐user concern with pre‐use testing is that the additional equipment needed for the test would be complex, have increased process risk, and unnecessarily increase cost. The European Medicines Agency’s motivation for this continued enforcement is twofold; one to address the

concern that some filter flaws might be masked during filtration by fluid contaminants out of the fluid stream, another is that sterilization could lead to filter distor­ tion [16]. Albeit the industry debate goes on, recent large industry survey for aseptic processes showed that 55% of respondents do perform PUPSIT testing moti­ vated by increased sterility assurance, regulatory com­ pliance, or financial risk mitigation [17]. Independent from the above, FDA asks end users to consider if it is necessary to add a second serial filter to the filtration setup, also referred to as redundant filtration. This con­ sideration should be documented and based on pro­ cess‐based risk assessment that incorporates factors such as: feed bioburden, filter‐retention validation, fil­ ter‐integrity history, processing time, processing cost, product value, and reprocessing potential [18]. Do note that for final sterilizing‐grade filtration, no rework is typically possible as the product would have been filled in small primary packaging containers, hence redun­ dant filtration setups are frequently used [17]. In the middle of these debates, end users seek for state‐of‐the‐art‐compliant solutions for safe, consist­ ent, and robust final filter wetting; integrity testing; and filtration. This has been addressed by Pall Biotech in providing automated process arrangements able to cope with the variety of end‐user filter wetting strate­ gies and integrity test types. To facilitate operational control of such complex single‐use redundant filtration setup with PUPSIT and to declutter the final filling environment, these will mostly be positioned outside of the filling environment. This layout is made possible given the confidence in disposable sterile connectors that allow to have the final filtration setup connected to a filling setup in a theoretically, uncontrolled environ­ ment. The prevailing surrounding environment will be determined by the type of filling machine installed. Driven by reduction in microbial and/or chemical con­ taminants, a similar trend toward closed processing is observed with barrier technologies becoming the norm for filling machines. Transfer of the sterile effluent within a closed disposable filtration system toward the filling manifold to be used inside of the closed filling machines asks for a way of aseptic transfer between these two. The systems that allow to transfer presteri­ lized material inside a filling environment are called rapid transfer port systems. The usage of closed SUSs in aseptic processing has led to the development of new testing methods to verify its leak freeness (see also Chapter  2) or integrity. In some instances, the tradi­ tionally used filter test equipment can offer additional testing options to test overall single‐use flow paths. The rapid transfer port systems consist of two separate parts, called the alpha part and the beta part (Figure 25.5). The principle of work is that by connecting the beta part

291

292

25  Single‐Use Technology for Formulation and Filling Applications

(a) Alpha-port

Beta-port A

Isolator

B

Beta-bag

(b)

A

B

A

B

Isolator

(c)

Isolator

Figure 25.5  Schematic concept of an Alpha/Beta‐Port system. (a) Approach, (b) locking, (c) opening.

to the alpha part, presterilized ends would become connected and contaminated ends would be isolated. The alpha part, also called alpha port, should be foreseen on the filling machine. As for the beta part, for single‐use applications it is available as a sealed bag containing the material to be transferred or a connector through which fluid will flow. Regardless of the design, each of them has an alpha connection interface and a fluid connection interface that allows fluid to enter the beta part. For multi‐head filling machines, it is common to use beta bag systems docking onto 190 mm alpha ports. This size is minimally required to allow the transfer of parts of, if not the entire single‐use assembly which can be com­ prised of filters, connectors, biocontainers, tubing, and filling needles. Although final filling machines (Figure  25.6) can be equipped with an entirely presterilized, ready‐to‐use

aseptic filling manifold, this benefit can, in practice, only be obtained for peristaltic dosing pump situations. Indeed, only part of the flow path can be foreseen in sin­ gle‐use format for the following dosing principles: mass flow filling, time pressure filling, and rotary piston fill­ ing. Subtle differences in dosing accuracy and effect on drug quality of each of these filling technologies makes that all of them are still in use. Manufacturing flexibility asks for multiproduct sites to be able to handle a variety of dosing technologies as well as primary packaging con­ tainers. Therefore, many final filling suppliers offer flex­ ible combination machines able to switch a machine to particular dosing or packaging needs which are particu­ larly successful for clinical drug filling machines or new capacity expansions for which multiproduct flexibility is becoming the norm (Table 25.1). Mostly, the peristaltic pump system will be part of these combination filling machines as it is the only option that eliminates cleaning of the flow path. If not impeded by historical drug filing, end users will typically consider this dosing technology as a candidate. The single‐use fill­ ing arrangement for peristaltic pump will typically involve: surge biocontainer, suction tubing, pump tub­ ing, dose tubing, and final filling needle. The surge biocontainer, occasionally referred to as header biocontainer, will continuously be fed by the inline filtration setup and contain a target fluid level. The amount of fluid in this biocontainer is typically a multi­ ple of the amount of fluid consumed in one full machine dose cycle. The fluid availability in this surge biocon­ tainer allows to fulfill its bubble trap functionality thereby avoiding air to enter the filling lines. Indeed, suc­ tion of air into filling lines can lead to filling volume devi­ ations consequently leading to container discard. Other than fluid availability, filling lines will typically distribute fluid from this location to the filling nozzles, this is ide­ ally done in an equal and consistent manner to avoid fill­ ing line variations. Down the filling lines, one will reach the peristaltic pumps that either have a single or double hose arrange­ ment. The latter requires two additional Y fittings per fill line to branch from the suction tube to the pump tubes and from these to the dosing tube leading to the filling needle. The single hose pump principle provides a sim­ plified setup with reduced leakage risks for disposables, yet some industry professionals challenge its claimed equivalent accuracy compared to double hose technol­ ogy. In any case, peristaltic pump heads must be com­ bined with their intended tubing to support the final filling machine vendor in reaching the end user’s accu­ racy objective. As for the end user, it is recommended to verify ease of operation, to qualify the material (en­durance, adsorption, extractables, etc.), and to evaluate the impact on drug quality and efficacy like for any dosing

25.7  Summary and Conclusions

Figure 25.6  Swissfillon isolator final filling machine equipment from Optima.

Table 25.1  Overview of today’s small to mid‐scale combination final filling machine systems enabling flexible manufacturing. Supplier

Filling machine

Bausch & Stroebel

Varyosys

Optima

MultiUse

Groninger

Flexpro

Bosch

FXS Combi

Dara

Combo

Vanrx

Robotic Isolated Filling

system. A typical drug‐quality impacting example would be the suitability of the tube material to pump operation. Indeed, it should be verified that for the intended inten­ sity and duration of use, particulate generation stays within acceptance limits. The final filling needle’s functionality allows for accu­ rate, drip free, and uninterrupted dosing of DP into the target container. To reach this, filling needles will typi­ cally plunge into the primary container to start filling close to the bottom of the primary container while gradually lifting to the top near the end of the dose cycle. Therefore, needle straightness during such oper­ ation is of the highest concern as to avoid any contact

with the container. The filling needles are traditionally made from stainless steel but in recent years reinforced high‐performance plastics have shown to be identical, if not more accurate and consistent, in target dosing volume [19]. For disposable technology, components should be ready‐to‐use, single‐use, comply with rele­ vant pharmacopeias, allow for sterilization through irradiation, and be cost‐effective. Only few needles, such as the Allegro™ filling needle, meet the require­ ment of truly single‐use and ready‐to‐use which is a prerequisite for integration into SuSs. When compar­ ing candidate filling needle performances, end users would need to assess the following aspects with their DP: filling accuracy, tendency to drip, tendency for DP dry out on the needle tip, and straightness when filling desired target container.

25.7 ­Summary and Conclusions Single‐use technologies are suited for enabling safe, economic, and flexible manufacturing even at the most critical formulation and final filling steps. In‐depth evaluation of the supplier approach for component selection, design creation, assembly validation, man­ ufacturing conditions, and supply chain robustness allows for necessary end‐user application and business risk assessment.

293

294

25  Single‐Use Technology for Formulation and Filling Applications

Nomenclature ADCs APIs CQAs DP DS

Antibody–drug conjugates Active pharmaceutical ingredients Critical quality attributes Drug product Drug substance

FDA ISO mAb PUPSIT SUSs

Food and Drug Administration International Organization for Standardization Monoclonal antibody Pre‐use post sterilization integrity test Single‐use systems

­References 1 von Wintzingerode, F. (2017). Biologics production:

impact of bioburden contaminations of non‐sterile process intermediates on patient safety and product quality. https://www.americanpharmaceuticalreview. com/Featured‐Articles/337286‐Biologics‐Production‐ Impact‐of‐Bioburden‐Contaminations‐of‐Non‐Sterile‐ Process‐Intermediates‐on‐Patient‐Safety‐and‐Product‐ Quality (accessed 20 September 2018). 2 Bishai, D. (2010). Presentation. World Vaccine Congress,. Washington D.C. (19–20 April). 3 Das, N. (2016). Commercializing high‐concentration mAbs. http://www.biopharminternational.com/ commercializing‐high‐concentration‐mabs (accessed 20 September 2018). 4 Srinivasan, C., Weight, A.K., Bussemer, T., and Klibanov, A.M. (2013). Non‐aqueous suspensions of antibodies are much less viscous than equally concentrated aqueous solutions. Pharm. Res. 30 (7): 1749–1757. 5 van Reis, R. and Zydney, A. (2007). Bioprocess membrane technology. J. Membr. Sci. 297 (1–2): 16–50. 6 Morar‐Mitrica, S., Adams, M.L., Crotts, G. et al. (2018). An intercompany perspective on biopharmaceutical drug product robustness studies. J. Pharm. Sci. 107: 529–542. 7 Liu, C. and Downey, W. (2017). Biopharma fill finish contract manufacturing market. https://www. contractpharma.com/issues/2016‐11‐01/view_features/ biopharma‐fill‐finish‐contract‐manufacturing‐market (accessed 20 September 2018). 8 Bio Process System Alliance. (2017). Design, control, and monitoring of single‐use systems for integrity assurance. http://bpsalliance.org/wp‐content/uploads/2017/07/3‐ INTEGRITY‐TASK‐Force‐2017.pdf (accessed 20 September 2018). 9 BioPlan Associates. (2017). 14th annual report and survey of biopharmaceutical manufacturing capacity and production. http://bioplanassociates.com/wp‐content/ uploads/2015/10/14th‐Annual‐Biomfg‐Report_TABLE‐ OF‐CONTENTS‐LR.pdf (accessed 20 September 2018). 10 Ding, W., Madsen, G., Mahajan, E. et al. (2014). Standardized extractables testing protocol for single use Systems in Biomanufacturing. Pharm. Eng. 34 (6): 1–11.

11 Gikanga, B. (2016). Processing impact on monoclonal

antibody drug products: protein subvisible particulate formation induced by grinding stress. PDA J. Pharm. Sci. Technol. 71: 172–188. 12 European Commission. (2008). EU GMP Annex 1 manufacture of sterile medicinal products. https://ec. europa.eu/health/sites/health/files/files/eudralex/ vol‐4/2008_11_25_gmp‐an1_en.pdf (accessed 20 September 2018). 13 Jameel, F., Hershenson, S., Khan, M.A., and Martin‐Moe, S. (2015). Quality by Design for Biopharmaceutical Drug Product Development. New York: Springer‐Verlag. 14 Greb, E. (2009). The debate over preuse filter‐ integrity testing. http://pharmtech.findpharma.com/ pharmtech/Article/The‐Debate‐over‐Preuse‐Filter‐ Integrity‐Testing/ArticleStandard/Article/ detail/612612 (accessed 23 December 2009 and 20 September 2018). 15 European Commission. (2017). EU GMP Annex 1 revision: manufacture of sterile medicinal products. https://www.gmp‐compliance.org/guidelines/gmp‐ guideline/eu‐gmp‐annex‐1‐revision‐manufacture‐of‐ sterile‐medicinal‐products‐draft (accessed 20 September 2018). 16 European Commission. (2007). European GMP guide annexes – supplementary requirements – Annex 1 manufacture of sterile medicinal products. Question (H+V June 2007): How should the integrity of sterilizing filters be verified? http://www.ema.europa. eu/ema/index.jsp?curl=pages/regulation/q_and_a/q_ and_a_detail_000027.jsp&mid=WC0b01ac05800296ca (accessed 20 September 2018). 17 Parenteral Drug Association. (2017). PDA aseptic processing survey. Question 146. https://store.pda.org/ ProductCatalog/Product.aspx?ID=4102 (accessed 20 September 2018). 18 Pharmaceutical Technology Editors. (2011). Can redundant filtration make sterility assurance double sure? http://www.pharmtech.com/can‐redundant‐ filtration‐make‐sterility‐assurance‐double‐sure (accessed 20 September 2018). 19 Zambaux, J.P. and Barry, J. (2014). Development of a single‐use filling needle. Bioprocess Int. 12: 46–53.

295

26 Facility Design Considerations for Mammalian Cell Culture Sue Walker Engineering Consultant, Portsmouth, NH, USA

26.1 ­Introduction During the design of a new facility or the renovation of an existing one, the initial design discussions are paramount to the success of the entire project. It is never too early for detailed planning as well as thorough documentation of all project‐related activities. Topics to be considered in these early discussions should focus on current needs as well as mid-range (one to five years) and long‐ term (five+ years) goals relating to all project objectives with a focus on business, technical, and regulatory. The business plan should include a market assessment both on a local and a global basis. Regulatory requirements is a broad category and can include individual corporate standards, industry regulations, regulatory standards and local, state, federal, and country laws. There can also be pharmaceutical standards, safety standards, and environmental standards. Regulatory requirements can be influenced by the markets defined in the business plan as they relate to compliance with Good Manufacturing Practices (GMP), local standards, and international regulations. The primary technical goal is to manufacture a safe, efficacious product. And the facility should always be designed with the end users in mind to provide a safe, practical and ergonomic work environment. Other technical goals can focus on process efficiency, equipment flexibility, facility organization, and overall economics such as capital expenses, operating expenses and cost of goods. This generic case study will focus on the process and the facility in the early design phases.

26.2 ­Generic Case Study 26.2.1  Generation of the Process Model The process model is a key deliverable in the early facility design project phases. The process model will require critical input from the business case such as product

type, number of products, projected amount of material required for each clinical phase, and market demands during commercial production. The business case may also consider the impact on return on investment between acquiring internal manufacturing capabilities vs. subcontracting of production to external contract manufacturing organizations. Contract development and manufacturing organizations are playing a strong role in this discussion with offerings from cell line development through all phases of manufacturing as well as collaborative agreements for manufacturing, equipment, and other required services. The importance of generating a process model as early as possible in the design of the facility cannot be overstated. If at all possible, the facility should be designed around the process and not the other way around. At a minimum, a process map is required to start the facility layout. This is a beneficial tool for tracking the project history by creating a clear, strong documentation trail as well as an early record of information required to generate any user‐requirements specification and to serve as the basis for the detailed design. Validation activities should also be considered and incorporated in these early phases. An example of a generic monoclonal antibody (mAb) process map is included in Figure 26.1. While each mAb project will be unique, at this stage it can be a reasonable assumption to follow a generic mAb process template with associated equipment if specifics are not known. Reasonable assumptions can also be made based on current industry practice and any internal expertise to generate a generic process map as changes will be made as knowledge is gained. It is important to note that the operating mode may not be known but it is often best to first assess your needs unit operation by unit operation. Regardless of the choice of operating mode, the process will be governed by the same parameters but the location of the operating window may change. And updating of

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

296

26  Facility Design Considerations for Mammalian Cell Culture

Seed expansion

CEX Aggregate, leached protein A and HCP reduction

CEX HCP reduction to meet specifications and enable ≥ 4 LVR viral clearance

Cell culture bioreactor

VI Enable ≥ 4 LRV viral clearance

NF Enable ≥ 4 LRV viral clearance

Harvest and depth filtration insoluble impurity removal

Capture (Protein A) HCP and DNA reduction and concentration

UF/DF Buffer exchange and concentration

Figure 26.1  Generic mAb process map.

the overall model will be an ongoing, dynamic process for the duration of the project. Once the process map is established, technology selection can begin. Single‐use technology is a good starting point because it “is now a complementary and compatible option, supported by cost‐effective and quick to build facilities, giving biopharma greater efficiency and flexibility in managing its capacity utilization” [1]. As equipment selections are made, they should be used to populate and to expand the process map which will ultimately lead to the creation of the process flow diagram(s). Once the equipment selection stage is reached, it will be dependent on the mode of operation whether batch/fed‐ batch vs. connected vs. continuous. The term connected is being used to describe a process that incorporates both continuous unit operation(s) and traditional batch unit operation(s). The term continuous refers to an end‐to‐ end operation with continuous flow for extended periods of time and minimal holdup volume. Although continuous technology has been used successfully in other industries, it is evolving now for the biopharmaceutical industry with encouragement from regulatory authorities. There are no general guidelines as to which mode to select as each molecule and each project is unique and evaluation of several different scenarios may be beneficial. Batch and connected may be less risky as there is more time to respond to an adverse event and “end‐to‐ end continuous bioprocessing faces some implementation challenges. These challenges center around three themes: regulatory, expertise/complexity and control strategy” [2]. A paper presented by Pollock et  al. [3] ­provides an interesting reference as the authors assessed the feasibility of continuous bioprocessing for the full ­product life cycle of a mAb. There are also a growing number of published studies to reference on this topic.

But, if internal capabilities are not available to make a unique project assessment, there are a number of equipment vendors, service providers, and software applications that may be of assistance. While mode of operation is one, equipment selection can be impacted by many factors including: ●● ●● ●● ●● ●● ●● ●● ●● ●● ●● ●● ●● ●● ●● ●● ●● ●●

Type and number of products Amount of product to be produced Type of process Scale of operation Any existing process platforms Number of operating shifts Level of automation Level of required operator training Vendor (price, service, and reputation) Cleanability Calibration and maintenance needs Space constraints Previous experience (both positive and negative) Equipment standardization GMP and other regulatory requirements Operator ease of use Safety and ergonomics

As each unit operation and its associated equipment are being refined, it is time to start to give some thought to what rooms they will occupy in the facility. For example, the mAb upstream processing (USP) suite can include inoculation, seed expansion, bioreactor run, and clarification. Inoculation and seed expansion may occur in a room with equipment such as a biological safety cabinet (BSC) or isolator system, an incubator and a small bench containing general supplies, and/or instrumentation such as for cell counting. Seed expansion can be efficiently carried out in flask(s) and/or wave‐mixed bioreactor(s).

26.2  Generic Case Study

A wave‐mixed bioreactor is economical and easily transportable (up to 20/50) making it a good choice to seed the smaller production bioreactors. The USP suite may contain production‐scale bioreactor(s), controller(s), clarification skid(s), floor scale(s), single‐mix vessel for the clarified material, bioprocess containers and bench(s) for in‐process control (IPC) measurements, required raw materials, and/or general operator use. Single‐use bioreactors (50, 200, 500, 1000, and 2000 l) and clarification by depth filtration are the workhorses of the mAb USP. The mAb downstream processing (DSP) suite may need to be contained in two separate areas – a pre‐viral suite(s) and a post‐viral suite. The traditional mAb purification process is two to three chromatography steps that are carried out in the pre‐viral suite. Equipment needs include at least one chromatography skid and one nanofiltration skid as well as mix vessels (for viral inactivation and intermediate collection), bioprocess containers, floor scale(s), and benches. If traditional columns are being used, planning for column packing whether in the DSP suite or in a separate, dedicated room should be considered. If only one chromatography skid is used with traditional columns, an assessment of all process steps (wash, elution, sanitization, regeneration, storage, etc.) and their impact on throughput should be conducted. The post‐ viral suite can include a tangential flow filtration skid as well as mix vessel, bioprocess containers, floor scale(s), and benches. Inline buffer dilution or inline buffer conditioning skid(s) may also be included in the DSP suite(s). Media and buffer (solution) preparation is a critical function for day‐to‐day plant operations and can require significant volumes of water as well as have a significant impact on storage space. Inline dilution and/or inline conditioning can be a good option to increase capacity and to decrease required space. Unless the strategy has been set, it is best to perform an evaluation at the worst‐ case (1×) condition to understand the space required to make, to handle, and to store all containers at the appropriate conditions as well as the impact on water requirements (especially, if water for injection will be used for the majority of the buffers). The estimation of the required media volumes is fairly straightforward at approximately 1–1.3× the total bioreactor volumes. Buffer volumes can range from 3000 to 5000 l per batch at the 1000 l scale and 8 500–13 000 l at the 2  ×  2000 l scale. These are estimates for batch/fed‐batch operation only and they may also change with titer, process template, etc. One source states that there is a “5–10‐fold increased rate of raw material ­consumption under continuous upstream processing (perfusion‐media consumption)” [4]. Regardless of the method of solution preparation or the overall mode of process operation, solution preparation is a serious contender for utilities and space utilization including the space for moving and storing buffers even with just in time preparation.

The handling of the bulk drug substance (BDS) and any final filling operations (compounding, final filtration, and filling) will not be discussed in detail but again will have a significant impact on the overall facility design and operation. Storage conditions for BDS may be refrigerated or frozen requiring properly sized, contained, and monitored storage units as well as accommodations for proper thawing. The compounding activity may require segregation from the final filtration and filling line. Isolators, robotics and other flexible filling solutions and technologies are worth investigating if filling capabilities are required. A generic example of a 1000 l mAb process flow diagram is included in Figure 26.2. It is provided for illustrative purposes only and may change significantly with incoming titer. All buffer volumes are order of magnitude estimates and can be either 1X or concentrates. The more details that are included from the beginning, the better the required space estimate and the greater the chance that all functionality will be designed in. Assumptions are allowed as it is expected that this will be a dynamic, learning process. The overall goal of this activity is to generate a process model that includes specific information for each step or unit operation such as material balances, all required equipment, all process parameters and their critical set points, all media, feed and buffer requirements, etc. Although the process model will remain dynamic over the course of the project, the time investment in the process map and the process flow diagram (and subsequently the process model) in these early project stages will significantly impact future success. 26.2.2  Generation of the Facility Description There are many factors to consider during the initial facility design phases. The top two are GMP compliance (as well as all required standards, regulations, and laws) and operator ease of use. There should be unambiguous definition of GMP zones to minimize risk of contamination and cross contamination with clear personnel, material, and waste flows. These flows should be unidirectional, if possible or as required. Materials management and optimization of materials handling can be a tremendous time and energy saver. Optimization can have a dramatic impact on overall operational efficiency both current capacity and future expansion. Operator ease of use is straightforward. The facility should be designed with the day‐to‐day and even hour‐by‐hour activities of each operator in mind. The creation of a safe, ergonomic, and common sense work place can pay long‐ term dividends in employee satisfaction and retention. Budget and return on investment are also critical for long‐term business success. Budget will play a strong role if evaluating the trade‐offs between capital investment and timing for traditional vs. modular build.

297

Seed medium Seed medium Seed medium Production medium Feeds (up to 4) Buffer

1l 10 l 150 l

Flask 50 l Wave 200 l SUB

900 l 300 l

1000 l SUB

600 l

Depth filtration

Control: T, DO, O2 Monitor: S, pH, W

0.2 μm filtration 2500 l SUM Buffers (up to 5) Post production buffers Buffers (up to 2)

600 l 600 l 25 l

Control: T Monitor: T, S, W

Capture: 1 cycle D = 80 cm H = 20 cm VI – 200 l SUM

Monitor: S, pH, W

0.2 μm filtration

Buffer

200 l 350 l

Buffers (up to 2) Post production buffers

350 l

500 l SUM CEX : 1 cycle D = 60 cm H = 20 cm 0.2 μm filtration

Buffer

25 l

Buffers (up to 4)

600 l

Post production buffers

400 l

2000 l SUM

Control: T Monitor: T, S, W, pH

AEX : 1 cycle D = 60 cm H = 20 cm 0.2 μm filtration 500 l SUM

Buffer

100 l

NF

0.2 μm filtration 500 l SUM Buffer

200 l

Non-production buffers

300 l

Buffer

30 l

TFF

200 l SUM

Monitor: S, W

0.2 μm filtration 200 l SUM BDS

Figure 26.2  Generic mAb process flow diagram.

Monitor: S, W

26.2  Generic Case Study

Area

Priority

Class

Function

Size (sq. ft.)

Major equipment

A

C

Seed expansion

500

Incubator, BSC, WAVE

A

D

USP cell culture and harvest

750

Seed SUB, SUB, clarification skid

A

C

Pre-viral area capture, VI, CEX, AEX, NF

1000

Chromatography skid (×2), SUM (×4) columns (×3), VF skid

A

C

Post-viral area UF/DF

500

TFF skid, BSC, SUM

A

D

Media preparation

500

Chemical hood, weighing station, SUM

A

D

Buffer preparation

500

Chemical hood, weighing station, SUM

B

D

Column packing

500

Floor scale, packing skid

A

CNC

Utilities

400

RO/DI, WFI, Clean steam

B

CNC

Warehouse

2500

Shelving system

C

CNC

Labs

1500

Based on function

Manufacturing

Preparation

Support

Figure 26.3  Example of a master “wish” list.

One of the most basic ways to start the evaluation of the facility (or more simply a building) is by the number of floors. A single‐floor building is the cheapest and the easiest to build or lease but ceiling heights may be an issue. Mezzanines usage may be an option but always consider floor loading. In general, always consider slab loading as a basic commercial building may not be able to accommodate the weight of even a 500 l bioreactor. A two‐story building is more expensive to build or to lease but has more vertical height options. The segregation of different areas may be easier but also may create complexity such as in raw material and intermediate/product transport or waste removal. A multistory building is the most expensive but there is more available space and more flexibility in custom configurations. Space allocation can be more complex, so it is critical to evaluate the overall flow and manufacturing layout. Height should no longer be an issue. A simple way for the preliminary assessment of the area required for a new space or building or the preliminary space allocation for an existing area is the creation of a master “wish” list. This can be a spreadsheet containing all rooms or suites under consideration, room classifications, estimates for intended areas, large equipment requiring floor space, and any known prioritization in the event that later cuts need to be made. Areas can be

estimated based on any experience as well as an assessment of routine equipment footprint with the addition of a safety margin for personnel movement and general clearance (Figure 26.3). Although it is recognized that air locks are required, the amount of space they require can be underestimated. Airlocks will be required for personnel, equipment, waste and product movement, and depending on applicable regulations, these streams may need to be separated. Airlocks may need to be sized to accommodate multiple people and gowning activities. The minimum size allowed for a basic airlock should be approximately 5–7 sq.m. (54–75 sq.ft.). 26.2.2.1  Manufacturing Areas

The following has been provided to aid with the creation of the manufacturing‐area section of a master list but it should not be considered fully inclusive. Depending on the functions required in the facility, the manufacturing areas may contain some or all of the following rooms. In some instances, room classifications have been suggested but consultation and compliance with all applicable project‐specific regulations is required as well as harmony with the proposed facility layout. References to room classifications assumes the following conversion: Class D equivalent to  International Organization for Standardization

299

300

26  Facility Design Considerations for Mammalian Cell Culture

(ISO) class 8, Class C equivalent to ISO class 7, and Class A equivalent to ISO class 5. USP Cell banking room – Class C Inoculation and cell expansion room Class C with a BSC or isolator (Class A) Upstream suite – Class D (bioreactor is considered as a closed system) DSP Pre‐viral downstream suite – Class D or Class C Post‐viral downstream suite – Class C (certain circumstances such as closed or contained processing may allow for one DSP suite or lower classification) Column packing room – Class D Solution Preparation Buffer and/or media preparation suite – Class D (separate rooms and/or special gowning may be required) Buffer and/or media storage at ambient and/or refrigerated conditions (in‐suite or in separate Class D storage room and to accommodate appropriate volumes) BDS and Filling BDS holding room  –  Controlled, nonclassified (CNC) with appropriate equipment to maintain storage temperature BDS compounding room – Class C Final filling area – Class C with Class A isolation technology for filling operation Final product inspection area – CNC Common Rooms in Manufacturing Area Equipment and parts preparation area and clean storage area IPC room (or in‐suite) General tank or bioprocess container staging area Janitor room – (A well thought‐out cleaning strategy can eliminate future problems from both contamination and cross contamination.) 26.2.2.2  Support Areas

The following has been provided to aid with the creation of the support area section of a master list but it should not be considered fully inclusive. Depending on the functions required in the facility, the support areas may contain some or all of the following rooms. General Possible significant space contributors include rooms for changing into the plant uniform and/or gowning, toilets, showers, and lockers as well as desk/office and conference room space. A security station, control room, and computer server room may also fall into this category.

Warehouse and Ancillary Functions The general function of the warehouse is an area that receives, stores, and ships materials for the facility. These materials can be raw materials required for the process and the general operation of the facility or any equipment. They can also be any products (BDS or drug product, DP) produced in the facility or the removal of any waste created in the facility. Proper segregation, handling, and storage of these different types of materials may be required in the warehouse or adjacent areas. If a significant number of single‐use assemblies are used in the facility, it may increase the size of the warehouse both in terms of handling and storage. One way to avoid excess handling is to quarantine and to release electronically in place. Only rejected materials would be moved to dedicated, segregated controlled areas. Inventory control and quality control (QC) may also require space in either the warehouse or an adjacent separate area for sampling, testing, and retain storage. Labs Process development and QC functions may require significant amounts of space depending on their scope of operation. QC testing such as microbiological testing, chemical analysis, analytical chemistry, and molecular biological testing such as polymerase chain reaction may require separate rooms. Separate analytical labs may also be required for the testing used to support development activities vs. the testing required to support GMP activities. Utility and Mechanical The utility and mechanical areas of the facility should be allocated space and designed with the same criteria as the manufacturing areas of the facility. This will provide long‐term benefits for day‐to‐day operation as well as ease of ongoing preventative maintenance. Based on process and facility specifics, there can be many critical systems that fall into this category but this discussion will be limited to general comments on autoclaves and heating ventilation air conditioning (HVAC) systems. Autoclave(s) The autoclave may be a surprising addition to a single‐ use facility but before the elimination of the preparation autoclave, careful consideration needs to be given to all materials and equipment required for all functions. Simple pieces of tubing and connectors, spatulas, filters, sampling devices, bioreactor loops, etc. may require autoclave treatment. Additionally, an autoclave may be required for waste inactivation as well as for an on‐ site microbiology lab. An appropriate water source will also be required to supply the autoclave.

­  References

HVAC [5] The energy required for HVAC to control the process environment consumes more energy than any other system type from 35 to 40% for BDS and 60 to 75% for DP. The simplest way to reduce the demand on HVAC is to minimize the quantity of air it handles. Air‐change rate should meet room classification but it does not have to be higher. A second way is to minimize the energy input into the conditioning of the air. This can be accomplished by turning off when not in use or by adjusting the set points to allow the system to run at peak efficiency. Optimized system performance and appropriate routine maintenance is critical to ongoing efficiency efforts.

26.3 ­Summary and Conclusions The ultimate goal is to have flexibility to go from pilot scale to clinical trials to commercial production in a fast and efficient manner. The process and the facility

considerations to accomplish this goal can include product range, scale of operation, economics, best practices, organizational structure, technology, stan­ dardization and regulatory aspects, to name just a few. But all requirements can be successfully met with careful planning and upfront evaluation of the process model, the facility layout, the general facility organizational pro­cedures, the hygiene concept, and the utility concept. The initial project stages should be dynamic and forward‐looking and supported by sound technical decision‐making. Project alternatives should be considered and documented to meet the business model as well as the process model and the facility requirements. The end‐user experience should always be a top consideration. The power in the early design stages can result in lower overall cost and early resolution of critical issues with the possibility to develop alternatives if changes are required in later stages. It can also expedite subsequent stages while still increasing accuracy.

Nomenclature BDS BSC CNC DSP DP GMP

Bulk drug substance Biological safety cabinet Controlled, nonclassified Downstream processing Drug product Good Manufacturing Practices

HVAC IPC ISO mAb QC USP

Heating ventilation air conditioning In‐process control International Organization for Standardization Monoclonal antibody Quality control Upstream processing

­References 1 Stanton, D. (2017). Modular single‐use facilities

rebalance biopharma’s capacity scales. https://www. biopharma‐reporter.com/Article/2017/08/02/Modular‐ single‐use‐facilities‐rebalance‐Biopharma‐s‐capacity‐ scales (accessed 1 August 2018). 2 Galliher, P., Jagschies, G., and Dua, A.R. (2017). Continuous bioprocessing: is it for everyone? https:// www.genengnews.com/gen‐articles/supplement‐ continuous‐bioprocessing‐is‐it‐for‐everyone/6146 (accessed 1 August 2018).

3 Pollock, J., Coffman, J., Ho, S.V., and Farid, S.S. (2017).

Integrated continuous bioprocessing: economic, operational, and environmental feasibility for clinical and commercial antibody manufacture. Biotechnol. Prog. 33 (4): 854–866. Challener, C.A. (2018). Managing uncertainty in 4 continuous biomanufacturing. BioPharm. Int. 31 (5): 12–17. Markarian, J. (2017). Designing sustainable pharma 5 facilities. Pharm. Technol. 41 (12): 48–49.

301

303

27 Progress in the Development of Single‐Use Solutions in Antibody–Drug Conjugate (ADC) Manufacturing Diego R. Schmidhalter, Stephan Elzner, and Romeo Schmid Lonza AG, Visp, Switzerland

27.1 ­Introduction The industry embraced single‐use technology (SUT) soon after the launch of the first disposable production solutions. Monoclonal antibodies (mAb) represent an important product class from a market point of view, and the platforms for production of these entities are characterized by a high level of uniformity. Therefore, equipment suppliers initially focused on the development of equipment and related single‐use components for the manufacture of this product group. In the last decade, antibody–drug conjugates (ADCs), in a broader sense also referred to as bioconjugates, adopted utilization of single‐use manufacturing technology. This trend was driven by product safety considerations (avoidance of cross contamination) and occupational exposure limit (OEL) requirements. In spite of initial ­setbacks in the clinic, ADCs represent a growing market segment with over 70 products in clinical development and 4 products launched, i.e. Kadcyla™ (Roche‐Genentech), Adcetris™ (Seattle Genetics), Mylotarg™, and Besponsa™ (Pfizer) [1]. From a mode of action point of view, ADCs are considered to be chemotherapeutics, as they combine the cytotoxic or cytostatic activity of a small molecule with the selectivity of a mAb. The antibody unit recognizes a tumor antigen specifically and, as a result, the ­payload can be delivered to the desired target. The systemic toxicity of these payloads is far too high for use without a targeting function. A key production step is the formation of a covalent chemical bond between the mAb and the highly potent payload, i.e. the active pharmaceutical ingredient (HPAPI), also called “warhead” or “toxin”. Common payloads are maytansinoids (DM1 and DM4), auristatins (MMAE and MMAF), and pyrrolobenzodiazepine (PBD) derivatives, with the latter showing the highest potency. Payloads are conjugated to the mAb through linkers of varying complexity. In

most cases, these linkers bind to either cysteine or lysine residues in the constant domain of a mAb. Cysteine‐based and lysine‐based conjugations [2] are shown schematically in Figure 27.1. For more detailed information on structural and functional aspects of ADCs, we refer interested readers to the review paper of Hoffmann et al. [1]. In addition to the abovementioned ADCs, the term bioconjugates includes other chemically modified proteins which are typically not classified as HPAPIs such as: ●●

●●

●●

●●

polymer (e.g. polyethylene glycol, hydroxyethyl starch, and polysialic acid)‐conjugated mAbs, in which the function of the polymer unit is to alter the pharmacokinetics of the mAb [3] mAbs conjugated to antibiotics for efficacious targeted treatment of bacterial infections [4] mAbs conjugated with a chelating agent which binds radionuclides, for delivery of cytotoxic radiation to target cancer cells (radioimmunotherapy) [5] and a conjugate vaccine of a weak antigen (e.g. a bacterial polysaccharide) with a strong antigen of protein nature, which elicits a stronger immunological response against the weak antigen [6].

Though selectivity is typically ensured through the use of a full mAb, this function could be taken over by an engineered antibody such as fragment antigen binding (Fab), a single‐chain variable fragment (scFv), a single‐ domain antibody (sdAb), or another antibody mimetic scaffold (affibody, DARPin, anticalin, alphabody, etc.) [7]. The structural complexity of bioconjugates, combined with the high potency in the case of ADCs, makes manufacture of these products a challenging task. In 2007, Lonza was one of the first contract manufacturing organizations to invest in dedicated bioconjugation facilities, and the focus at that time was on ADCs. In

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

304

27  Progress in the Development of Single‐Use Solutions in Antibody–Drug Conjugate (ADC) Manufacturing Antibody modification

Toxin conjugation

Reducing agent SH

mAb

Linker-drug S

n

Linker

Drug n

Antibody–drug conjugate

Modified mAb

DAR-defining step

Reagent

Reagent

Partial reduction

Reducing reagent (e.g TCEP)

Drug derivative (e.g vcMMAE)

Antibody modification

Linker

Toxin conjugation

H N

Drug Linker

H N

Linker

Drug

n

mAb

n

Antibody–drug conjugate

Modified mAb

DAR-defining step

Reagent

Reagent

Lysine modification

Activated linker (e.g. SMCC)

Drug derivative (e.g DM1)

Figure 27.1  Typical antibody modification followed by conjugation for cysteine-based (above) and lysine-based (below) conjugation processes. Figure reprinted by permission from Springer Nature: Humana Press, copyright [2].

the 10 years after commissioning, more than 500 clinical and commercial batches from more than 30 different products were released from this facility. Glass and stainless‐steel multiuse equipment was used for initial production and adoption of disposable technology was limited to sterile filters, transfer tubing, sampling devices, and storage bags. In 2014, a fully disposable single‐use production platform was introduced which comprised 50–200 l single‐use mixing bag technology and tangential flow filtration (TFF) skids suitable for a membrane area of up to 3 m2. Since then, the majority of clinical programs were executed in end‐to‐end SUT production setups at even larger scale. This development was driven by benefits such as: ●●

●●

●●

●●

reduced risk of exposure to toxic compounds for the operators avoidance of the requirement for dedicated equipment for production of each individual HPAPI, which is a significant drawback when program turnover is high reduced implementation time in production due to elimination of cross‐contamination concerns; cleaning programs, which start with a cleaning risk assessment, and extend to execution of coupon studies, development of product‐specific cleaning procedures and cleaning validation, have become obsolete reduced analytical effort for cleaning monitoring taking into account that the steadily increasing potency of ADCs is challenging the detection and quantification limits of the available analytical methods and

●●

reduced quantities of toxic waste streams from the cleaning process.

Drawbacks of the use of polymer‐based production equipment are discussed in the following section of this chapter describing the progress in the development of single‐use solutions in ADC manufacturing.

27.2 ­Challenges for the Use of Disposables in ADC Processes 27.2.1  Use of Organic Solvents One challenge is the exposure of product‐contacting surfaces to organic solvents, which can have a negative  impact on physical stability and can give rise to  increased levels of leachables. Food and Drug Administration and European Medicines Agency regulations specifically state that processing equipment including single‐use systems should not present any hazard to product safety and quality. The typically lipophilic payload is usually dissolved in an organic ­solvent, and polar aprotic solvents such as dimethyl sulfoxide, dimethylformamide, or dimethylacetamide ­ are commonly used. In addition, the chemical reaction is usually performed in buffered aqueous solutions containing 10–25% organic solvent in order to maintain the solubility of the small molecule during the conjugation step. The subsequent purification steps should be designed to purge solvents, leachables, conjugation

27.2  Challenges for the Use of Disposables in ADC Processes

r­eagents (i.e.  activating agents), uncoupled payload, and quenching agents. The single‐use materials which come into contact with process solutions should be carefully assessed for compatibility under the chosen solvent concentrations and process conditions. Weibing Ding [8] has noted the following aspects which are to be  considered in a science‐based risk assessment: (i) chemical compatibility between the process fluid and single‐use systems, (ii) product composition, (iii) material contact area, (iv) contact time, (v) process temperature, (vi) presterilization method, and (vii) proximity to the final container closure system. It is fortunate that product contact layers of most bioprocess containers and single‐use mixing bags consist of polyethylene which is compatible with aqueous solutions containing less than 25% solvent. However, single‐use components such as connectors made of polysulfone, polyethersulfone, polycarbonate, or polyvinylidene fluoride (see also Chapter  5) may have lower resistance to  similar water–solvent mixtures, which can lead to increased leaching of chemicals, thus impacting product quality. In addition, the problems associated with a solvent content of 25–100% pose a challenge, which can be met by use of glass bottles and stainless‐steel connectors in single‐use mode. In summary, the combination of a thorough leachables risk assessment (see also Chapter 11) and careful selection of single‐use equipment components are the activities which enable reduction of the concentrations of leachables in the final product to an acceptable level.

27.2.2  Safety and Handling of HPAPIs ADC production requires handling of highly potent components with cytostatic or cytotoxic characteristics. Occupational safety and health, and the avoidance of cross contamination therefore require specific attention. Protection of employees or contract workers from any negative impact on health due to exposure to chemicals handled during manufacturing is therefore of the utmost importance. The chemical industry has developed a ­classification system for chemical compounds, including drug intermediates and drugs, which is based on ­defining OELs, which are the maximum acceptable concentrations of defined chemicals in the air of the working area. Suggested classification systems have evolved over time  into the SafeBridge standard [9], which has four categories. Typical OELs for chemicals used in ADC pro­duction range between 100 and 1 ng/m3. Table 27.1 summarizes information and data related to the four SafeBridge categories. The holy grail with regard to worker protection and occupational health is “contained manufacturing”, and stainless‐steel equipment interconnected with fixed stainless‐steel piping is a classical solution for contained production. In recent years, we have, however, been able to demonstrate that commercially available single‐use equipment is as good a choice from an occupational safety and cross‐contamination control point of view, and it can be concluded that single‐use process solutions are highly suitable for closed, aseptic processing. The chosen solutions work reliably and are able to withstand

Table 27.1  SafeBridge product classification categories 1–4. SafeBridge band

Category 1

Category 2

Category 3

Category 4

OEL

Over 500 μg/m3

500–10 μg/m3

10 μg/m3 to 30 ng/m3

Below 30 ng/m3

Toxicity and potency

Low

Moderate

Potent

Highly potent

Typical dosage (mg/kg)

>10

1–10

0.01–1

14

1 433 5 576

n.a.

[47]

ZRP Meander bioreactor

>20

5000

2 × 109

Own data, not published

Well plates

14–24

n.a.

n.a. (infusion: ≤1010)

[48] [49]

TIL

CAR T cells

NK cells

MSC

Bag

12

10.6

n.a.

WAVE

35

n.a.

109

Bag

≤48

15 000

~3 × 10

VueLife bag system

14

80–200

n.a.

[50] 11

[51] [52]

T‐flaks

10

40

n.a.

[53]

VueLife bag system

21

277

n.a.

[54]

WAVE

21

12–354

n.a.

[55]

G‐Rex flask

8–10

442

n.a.

[56]

ZRP type M

30

1000–2000

n.a.

Own data, not published

ZRP type M

30–35

1000–50 000

n.a.

Own data, not published

Flasks

30

n.a.

n.a.

[57]

Flasks

30–45

6–52

n.a.

[58]

5‐Layer flasks

10–28

n.a.

n.a.

[59]

5‐Layer flasks

28

5–145

n.a.

[60]

5‐Layer flasks

22–28

n.a.

n.a.

[59]

n.a. = not available. Source: © Ralf Pörtner, Shreemanta K. Parida, Christiane Schaffer, Hans Hoffmeister. Originally published in [1] under CC BY 3.0 license. Available from: https://www.intechopen.com/books/stem‐cells‐in‐clinical‐practice‐and‐tissue‐engineering/ landscape‐of‐manufacturing‐process‐of‐atmp‐cell‐therapy‐products‐for‐unmet‐clinical‐needs

deeper characterization of ex vivo expanded immune cells is urgently needed. This applies not only on the level of a few receptors and ligands on the cell surface but also  with respect to the ever‐contained subtypes in an  expanded immune cell population, the pattern of secreted effector molecules, and their amounts over time as well as influences from in vivo components on them. More research on the aspects of modern cell ther­ apy  might be qualified as too costly but will be more

targeted and will at least avoid expensive and unjusti­ fied clinical studies maximizing the best use of the available R&D resources for better outcomes. During the last years, the particular nature of living immune cells has become more and more visible, a more ­sophisticated biotechnology and its analytical meth­ ods are coming up and what is most important, very promising treatments of late‐stage tumor patients have been published.

331

332

30  Single‐Use Bioreactors for Manufacturing of Immune Cell Therapeutics

Nomenclature T cells TIL CAR‐T cells NK cells MSC

T lymphocytes Tumor‐infiltrating T cells Chimeric antigen receptor T cells Natural killer cells Mesenchymal stem/stromal cells

ATMP GMP BM CD

Advanced therapy medicinal products Good manufacturing practice Bone marrow Cluster of differentiation

­References 1 Pörtner, R., Shreemanta, K.P., Schaffer, C., and

Hoffmeister, H. (2018). Landscape of manufacturing process of ATMP cell therapy products for unmet clinical needs. In: Stem Cells in Clinical Practice and Tissue Engineering (ed. R. Sharma). Intechopen. doi: 10.5772/65995. 2 Dominici, M., Le Blanc, K., Mueller, I. et al. (2006). Minimal criteria for defining multipotent mesenchymal stromal cells. The International Society for Cellular Therapy position statement. Cytotherapy 8 (4): 315–317. 3 Smith, C., Økern, G., Rehan, S. et al. (2015). Ex vivo expansion of human T cells for adoptive immunotherapy using the novel Xeno‐free CTS immune cell serum replacement. Clin. Transl. Immunol. 4 (1): e31. https://doi.org/10.1038/ cti.2014.31. 4 Rosenberg, S.A., Yang, J.C., Sherry, R.M. et al. (2011). Durable complete responses in heavily pretreated patients with metastatic melanoma using T‐cell transfer immunotherapy. Clin. Cancer Res. 17 (13): 4550–4557. 5 Gattinoni, L., Klebanoff, C.A., Palmer, D.C. et al. (2005). Acquisition of full effector function in vitro paradoxically impairs the in vivo antitumor efficacy of adoptively transferred CD8+ T cells. J. Clin. Invest. 115 (6): 1616–1626. 6 Lanzavecchia, A. and Sallusto, F. (2001). Regulation of T cell immunity by dendritic cells. Cell 106 (3): 263–266. 7 Wu, R.C., Hwu, P., and Radvanyi, L.G. (2012). New insights on the role of CD8+CD57+T‐cells in cancer. OncoImmunology 1 (6): 954–956. 8 Kared, H., Martelli, S., Ng, T.P. et al. (2016). CD57 in human natural killer cells and T‐lymphocytes. Cancer Immunol. Immunother. 65 (4): 441–452. 9 Grossman, W.J. (2004). Differential expression of granzymes a and B in human cytotoxic lymphocyte subsets and T regulatory cells. Blood 104 (9): 2840–2848. 10 Almeida, J.R., Price, D.A., Papagno, L. et al. (2007). Superior control of HIV‐1 replication by CD8+ T cells is reflected by their avidity, polyfunctionality, and clonal turnover. J. Exp. Med. 204: 2473–2485.

11 Harari, A., Cellerai, C., Enders, F.B. et al. (2007).

12

13

14

15

16

17 18

19

20

21

Skewed association of polyfunctional antigen‐specific CD8 T cell populations with HLA‐B genotype. Proc. Natl. Acad. Sci. USA 104: 16233–16238. Wiesinger, M., Stoica, D., Roessner, S. et al. (2017). Good manufacturing practice‐compliant production and lot‐release of ex vivo expanded regulatory T cells as basis for treatment of patients with autoimmune and inflammatory disorders. Front. Immunol. 8: https://doi. org/10.3389/fimmu.2017.01371. Kishton, R.J., Sukumar, M., and Restifo, N.P. (2017). Metabolic regulation of T cell longevity and function in tumor immunotherapy. Cell. Metabol. 26 (1): 94–109. Hami, L.S., Green, C., Leshinsky, N. et al. (2004). GMP production and testing of Xcellerated T cells for the treatment of patients with CLL. Cytotherapy 6: 554–562. Duhen, T., Duhen, R., Montler, R. et al. (2018). Co‐expression of CD39 and CD103 identifies tumor‐ reactive CD8 T cells in human solid tumors. Nat. Commun. 9 (1): 2724. Powell, D.J. (2005). Transition of late‐stage effector T cells to CD27+ CD28+ tumor‐reactive effector memory T cells in humans after adoptive cell transfer therapy. Blood 105 (1): 241–250. Caligiuri, M.A. (2008). Human natural killer cells. Blood 112 (3): 61–69. Narni‐Mancinelli, E., Ugolini, S., and Vivier, E. (2013). Tuning the threshold of natural killer cell responses. Curr. Opin. Immunol. 25 (1): 53–58. Zamai, L., Ahmad, M., Bennett, I.M. et al. (1998). Natural killer (NK) cell‐mediated cytotoxicity: differential use of TRAIL and Fas ligand by immature and mature primary human NK cells. J. Exp. Med. 188 (12): 375–380. Screpanti, V., Wallin, R.P., Ljunggren, H.G., and Grandien, A. (2001). A central role for death receptor‐ mediated apoptosis in the rejection of tumors by NK cells. J. Immunol. 167 (4): 2068–2073. de Andrade, L.F., Smyth, M.J., and Martinet, L. (2014). DNAM‐1 control of natural killer cells functions through nectin and nectin‐like proteins. Immunol. Cell. Biol. 92 (3): 237–244.

­  References

22 Molfetta, R., Quatrini, L., Santoni, A., and Paolini, R.

(2017). Regulation of NKG2D‐dependent NK cell functions: the yin and the Yang of receptor endocytosis. Int. J. Mol. Sci. 18 (8): pii E1677. 23 Belting, L., Hömberg, N., Przewoznik, M. et al. (2015). Critical role of the NKG2D receptor for NK cell‐ mediated control and immune escape of B‐cell lymphoma. Eur. J. Immunol. 45 (9): 2593–2601. 24 Fionda, C., Soriani, A., Zingoni, A. et al. (2015). NKG2D and DNAM‐1 ligands: molecular targets for NK cell‐mediated immunotherapeutic intervention in multiple myeloma. Biomed. Res. Int. 2015: 178698. 25 Fernandez, L., Valentin, J., Zalacain, M. et al. (2015). Activated and expanded natural killer cells target osteosarcoma tumor initiating cells in an NKG2D‐ NKG2DL dependent manner. Cancer Lett. 368 (1): 54–63. 26 Pasero, C., Gravis, G., Granjeaud, S. et al. (2015). Highly effective NK cells are associated with good prognosis in patients with metastatic prostate cancer. Oncotarget 6 (16): 14360–14373. 27 Garcia‐Iglesias, T., Del Toro‐Arreola, A., Albarran‐ Somoza, B. et al. (2009). Low NKp30, NKp46 and NKG2D expression and reduced cytotoxic activity on NK cells in cervical cancer and precursor lesions. BMC Cancer 9: 186. 28 Schlegel, P., Ditthard, K., Lang, P. et al. (2015). NKG2D signaling leads to NK cell mediated lysis of childhood AML. J. Immunol. Res. 2015: 473175. 29 Pegram, H.J., Andrews, D.M., Smyth, M.J. et al. (2011). Activating and inhibitory receptors of natural killer cells. Immunol. Cell. Biol. 89 (2): 216–224. 30 Moretta, A., Bottino, C., Vitale, M. et al. (2001). Activating receptors and coreceptors involved in human natural killer cell‐mediated cytolysis. Annu. Rev. Immunol. 19: 197–223. 31 Wang, R., Jaw, J.J., Stutzman, N.C. et al. (2011). Natural killer cell‐produced IFN‐γ and TNF‐α induce target cell cytolysis through up‐regulation of ICAM‐1. J. Leukoc. Biol. 91 (2): 299–309. 32 Lettau, M., Schmidt, H., Kabelitz, D., and Janssen, O. (2007). Secretory lysosomes and their cargo in T and NK cells. Immunol. Lett. 108: 10–19. 33 Chowdhury, D. and Lieberman, J. (2008). Death by a thousand cuts: granzyme pathways of programmed cell death. Annu. Rev. Immunol. 26: 389–420. 34 Pipkin, M.E. and Lieberman, J. (2007). Delivering the kiss of death: progress on understanding how perforin works. Curr. Opin. Immunol. 19: 301–308. 35 Lieberman, J. (2003). The ABCs of granule‐mediated cytotoxicity: new weapons in the arsenal. Nat. Rev. Immunol. 3: 361–370. 36 Trapani, J.A. and Bird, P.I. (2008). A renaissance in understanding the multiple and diverse functions of granzymes? Immunity 29: 665–667.

37 Paolini, R., Bernardini, G., Molfetta, R., and Santoni, A.

38

39

40

41

42

43

44

45

46

47

48

49

(2015). NK cells and interferons. Cytokine Growth Factor Rev. 26 (2): 113–220. Smyth, M.J., Thia, K.Y., Cretney, E. et al. (1999). Perforin is a major contributor to NK cell control of tumor metastasis. J. Immunol. 162 (11): 6658–6662. Brower V. (2015). The CAR T‐Cell race: tumor‐ targeting T‐cell therapies are generating remarkable remissions in hard‐to‐beat cancers – and attracting millions of dollars of investment along the way. http:// www.the‐scientist.com/articles.view/articleNo/42462/ title (accessed 1 April 2015). Skrahin, A., Ahmed, R.K., Ferrara, G. et al. (2014). Autologous mesenchymal stromal cell infusion as adjunct treatment in patients with multidrug and extensively drug‐resistant tuberculosis: an open‐label phase 1 safety trial. Lancet Respir. Med. 2: 108–122. Subramani, B., Ratnavelu, K., Pullai, C.R. et al. (2013). Autologous immune enhancement therapy: a case report of a stage IV colonic cancer. Oncol. Lett. 5: 1611–1614. Chapuis, A.G., Roberts, I.M., Thompson, J.A. et al. (2016). T‐cell therapy using interleukin‐21‐primed cytotoxic T‐cell lymphocytes combined with cytotoxic T‐cell lymphocyte antigen‐4 blockade results in long‐term cell persistence and durable tumour regression. J. Clin. Oncol. 34 (31): 3787–3795. Itzhaki, O., Hovav, E., Ziporen, Y. et al. (2011). Establishment and large‐scale expansion of minimally cultured “young” tumour infiltrating lymphocytes for adoptive transfer therapy. J. Immunother. 34: 212–220. Sadeghi, A., Pauler, L., Annerén, C. et al. (2011). Large‐scale bioreactor expansion of tumour‐infiltrating lymphocytes. J. Immunolog. Meth. 364: 94–100. Jin, J., Sabatino, M., Somerville, R. et al. (2012). Simplified method of the growth of human tumour infiltrating lymphocytes in gas‐permeable flasks to numbers needed for patient treatment. J. Immunother. 35: 283–292. Somerville, R.P., Devillier, L., Parkhurst, M.R. et al. (2012). Clinical scale rapid expansion of lymphocytes for adoptive cell transfer therapy in the WAVE(R) bioreactor. J. Transl. Med. 10: 69. Donia, M., Larsen, S.M., Met, O., and Svane, I.M. (2014). Simplified protocol for clinical‐grade tumourinfiltrating lymphocyte manufacturing with use of the wave bioreactor. Cytotherapy 16: 1117–1120. Morgan, R.A., Yang, J.C., Kitano, M. et al. (2010). Case report of a serious adverse event following the administration of T cells transduced with a chimeric antigen receptor recognizing ERBB2. Mol. Ther. 18: 843–851. Tumaini, B., Lee, D.W., Lin, T. et al. (2013). Simplified process for the production of anti‐CD19‐CAR‐ engineered T cells. Cytotherapy 15: 1406–1415.

333

334

30  Single‐Use Bioreactors for Manufacturing of Immune Cell Therapeutics

50 Wang, X. and Riviere, I. (2015). Manufacture of

tumour‐ and virus‐specific T lymphocytes for adoptive cell therapies. Cancer Gene Ther. 22: 85–94. 51 Kershaw, M.H., Westwood, J.A., Parker, L.L. et al. (2006). A phase I study on adoptive immunotherapy using gene‐modified T cells for ovarian cancer. Clin. Cancer Res. 12: 6106–6115. 52 Luhm, J., Brand, J.M., Koritke, P. et al. (2002). Large‐ scale generation of natural killer lymphocytes for clinical application. J. Hematother. Stem Cell Res. 11: 651–657. 53 Torelli, G.F., Guarini, A., Maggio, R. et al. (2005). Expansion of natural killer cells with lytic activity against autologous blasts from adult and pediatric acute lymphoid leukemia patients in complete hematologic remission. Haematologica 90 (6): 785–792. 54 Fujisaki, H., Kakuda, H., Shimasaki, N. et al. (2009). Expansion of highly cytotoxic human natural killer cells for cancer cell therapy. Cancer Res. 69 (9): 4010–4017. 5 5 GE Healthcare. (2011). Application note 28‐9936‐25 AA. Perfusion culture of human natural killer cells in the WAVE Bioreactor 2/10 system. http://www. cellgenix.com/fileadmin/published_content/3_ downloads/4.1_application_notes/GE_paper_NK_ cells.pdf. 56 Lapteva, N., Durett, A.G., Sun, J. et al. (2019). Large‐ scale ex vivo expansion and characterization of natural killer cells for clinical applications. Cytotherapy 14 (9): 1131–1143. 57 Baron, F., Lechanteur, C., Willems, E. et al. (2010). Cotransplantation of mesenchymal stem cells might prevent death from graft‐versus‐host disease (GvHD) without abrogating graft‐versus‐tumour effects after HLA‐mismatched allogeneic transplantation following

58

59

60

61

62

63

64

65

66

nonmyeloablative conditioning. Biol. Blood Marrow Transplant. 16: 838–847. Emadedin, M., Ghorbani Liastani, M., Fazeli, R. et al. (2015). Long‐term follow‐up of intra‐articular injection of autologous mesenchymal stem cells in patients with knee, ankle, or hip osteoarthritis. Arch. Iran. Med. 18 (6): 336–344. Capelli, C., Pedrini, O., Valgardsdottir, R. et al. (2015). Clinical grade expansion of MSCs. Immunol. Lett. 168: 222–227. Lechanteur, C., Briquet, A., Giet, O. et al. (2016). Clinical‐scale expansion of mesenchymal stromal cells: a large banking experience. J. Transl. Med. 14 (1): 145. Diederichs, S., Röker, S., Marten, D. et al. (2009). Dynamic cultivation of human mesenchymal stem cells in a rotating bed bioreactor system based on the Z RP platform. Biotechnol. Prog. 25: 1762–1771. Lavrentieva, A., Hatlapatka, T., Neumann, A. et al. (2013). Potential for osteogenic and chondrogenic differentiation of MSC. Adv. Biochem. Eng. Biotechnol. 129: 73–88. Neumann, A., Lavrentieva, A., Heilkenbrinker, A. et al. (2014). Characterization and application of a disposable rotating bed bioreactor for mesenchymal stem cell expansion. Bioengineering 1: 231–245. Reichardt, A., Polchow, B., Shakibaei, M. et al. (2013). Large scale expansion of human umbilical cord cells in a rotating bed system bioreactor for cardiovascular tissue engineering applications. Open Biomed. Eng. J. 7: 50–61. Egger, D., Spitz, S., Fischer, M. et al. (2013). Application of a parallelizable perfusion bioreactor for physiologic 3D cell culture. Cells Tissues Organs. 203 (5): 316–326. Hoffmeister H. (2018). Patentschrift EP 2543719 A1.

335

Index a active pharmaceutical ingredient (API)  5, 287 ADCs. See antibody–drug conjugates (ADCs) adoptive T‐cell therapy  324 advanced therapy medicinal products (ATMP) therapy  329, 331 Allegro™ single‐use chromatography system 121 Allegro Single Use Mixer system  30 Allegro® STR system  45 Allegro® XRS 25 system  43 allogeneic engineered adoptive T‐cell therapeutics 324 alluvial filtration  273 American Chemical Society (ACS) 175 American Society for Testing and Materials (ASTM)  159 American Society of Mechanical Engineers‐BioProcessing Equipment (ASME‐BPE)  163, 165 analytical evaluation threshold (AET) 151 ancillary equipment  291 antibody–drug conjugates (ADCs) bulk drug substance freeze and thaw 308–310 chromatography 308 cysteine conjugation process  310–311 filtration and transfers  308 key unit operations  308–310 organic solvents uses  306–307 safety and handling of HPAPIs  307–308 stirred tanks reaction  308 aseptic connectors  57–59

aseptic coupling aseptic connectors  57–59 aseptic transfer systems  59–62 connection under laminar flow 57 steam‐in‐place connection  57 welding 59 aseptic disconnection  62–64 aseptic transfer device (ATD)  62 aseptic transfer systems  59–62 attenuated total reflection (ATR) technology 74 autoclave 302 autologous engineered adoptive T‐cell therapeutics  324 automation hardware systems  92 automation software  92

b Bacillus subtilis 38 Bacillus thuringiensis 75 bag chamber leak test (BC‐LT)  21–22 bag mixing systems classification 27–28 hydraulically driven  32–33 with oscillating devices  31–32 with rotating stirrer  28–31 with tumbling stirrer  31 bags bioreactors 38–40 chamber 22 for fluid and powder handling  13–15 fluid handling  14 for freezing and thawing  18 handling and container systems  15–18 manufacturing 110–113 powder handling  14–15

three‐dimensional (3D)  15 two‐dimensional (2D)  14–15 BarbLock system  54 biocompatibility testing  145 bioinformatics 84–86 biological safety cabinet (BSC)  298 BioPAT ViaMass Sensor  76 biopharmaceutical drugs, sterilization of 292 biopharmaceutical industry  181 cell and gene therapies  324–327 sustainability efforts  175–177 biopharmaceutical manufacturer’s approach chemical safety assessment  148–153 cost factor  154–155 life‐cycle management  155 production outsourcing and contract manufacturers  155 risk mitigation  146–148 supply chain of single‐use equipment 153–154 time factor  155 BioPhorum 222 BioPhorum Operations Group (BPOG)  148, 164, 165 BioPlan Annual Survey  198 BioProcessing Equipment (BPE) 163 Bioprocess Recirculating Mixing System 32–33 Bio‐Process Systems Alliance (BPSA)  104, 163–165, 177 bioreactors 195 biosensors 77–78 BioSolve Process  209 BIOSTAT® RM  43 BIOSTAT® STR  45

Single-Use Technology in Biopharmaceutical Manufacture, Second Edition. Edited by Regine Eibl and Dieter Eibl. © 2019 John Wiley & Sons, Inc. Published 2019 by John Wiley & Sons, Inc.

336

Index

biotechnology‐based drugs  83 blockbuster cancer drugs  83 Brevundimonas diminuta 128 bubble point (BP) test  138 bulk drug substance (BDS)  299

c Cadence™ BioSMB technology 282–284 cake filtration. See alluvial filtration cancer long‐known disease  322 stages 322 treatment evolution  323–324 capacitance sensors  75–76 caprolactam 318 carbon dioxide sensors  73 CAR‐T therapy  83 cell and gene therapies  321 adoptive T‐cell therapy  324 biopharmaceutical industry  324–327 cancer 322–324 nature 322 cell banking  314–315 cell‐based assay (CBA)  231 cell culture clarification model  274–275 CeLLine 40 CellMiscelatore from CerCell  30 cellulose acetate (CA)  128 cellulose nitrate (CN)  128 chemical compatibility tests  108 chemical safety assessment controlled extractables study 149 exposure scenario  153 extractables profiling  148 leachables study  152–153 simulated‐use extractables study  151–152 sum parameters  149–151 toxicological risk assessment  153 unknown compounds  151 Chinese domestic industry  196 Chinese hamster ovary (CHO) cell 232 cell harvest  275 cell lines  3 chromatography  120–121, 123–124, 308. See also continuous multicolumn chromatography cleanroom classification  226 cobalt‐60 (Co60)  223

Code of Federal Regulations (CFR)  86, 144, 145 Colder Products Company (CPC) 57 commercial manufacturing facilities 187 Committee on Environmental Improvement (CEI)  177 computational fluid dynamics (CFD)  237, 252, 269 computer‐aided design (CAD) software  237, 240 conductometric sensors  76 container closure integrity (CCI)  18–22 container systems  15–18 for in‐house applications  17 for liquid shipping  17–18 continuous chromatography capture 122 continuous downstream processing  282–284 continuous in‐process mixing and hold 123 continuous multicolumn chromatography 281–282 fed‐batch processes  282–284 perfusion processes  284 continuous processing  88–92 contract manufacturing organization (CMO)  195, 314 cooling temperature  234–235 cost‐effective manufacturing capabilities 205 implementing single‐use process platforms 206–209 stainless steel with single‐use process platforms  209–211 standardized single‐use process platforms 206 cost of goods (COGs)  205, 209, 210, 327 coupling aseptic coupling  57–62 aseptic disconnection  62–64 Cramer classification  151 crimping method  63 critical process parameters (CPPs)  233–235 critical quality attributes (CQAs)  231, 233, 287 cross flow  127 cumulative “worst‐case” models  214–215

current good manufacturing practice (cGMP)  83, 274, 313 Custom Single Run (CSR®) product line 29 cysteine conjugation process  310–311

d DAICEL CORPORATION  277 DAISEP MabXpure Technology  274 DECHEMA 164 depth filters  132–133 depth filtration  274 design of experiment (DOE)  234 diatomaceous earth (DE) filtration  273–274 diffusive flow (DF) test  138 diffusive mixing  25 direct flow/dead‐end filtration  127 dispersive mixing  25 disposable bags film manufacturing  99–102 Helium Integrity testing  115 materials 95–98 molding 102–104 principle 95 production 110–112 quality insurance  104–110, 112–113 disposable bioreactor systems  69–70 capacitance sensors  75–76 electrochemical sensors  76–77 interfaces 70–71 optical chemosensors  71–73 spectroscopic sensors  73–75 disposable membrane bioreactors  38–40 disposables  4, 6 distributed control system (DCS) 88 distributive mixing  25 dose maps  223–224 double‐door transfer system  61 downstream processing (DSP)  117, 184, 216, 299, 315–317 benefits and constraints  117 integrated continuous  124 intermediate‐scale single-use facilities 190–192 single‐use technologies  120–121 tools 90 trends 117–120 drug product (DP)  5

Index

drug substance (DS)  5, 287, 288, 315 drug‐to‐antibody ratio (DAR)  310 Dual Indicator System for Mixing Time (DISMT)  262, 263 dynamic mode  277

e electrochemical biosensors  77 electrochemical sensors categories 76 field‐effect transistors  76–77 pH electrode  76 Emerson Process Management DeltaV platform  87–88 end‐of‐life management  177–178 endotoxin test  113 environmental impacts, life cycle assessment 172–175 enzyme field‐effect transistors (EnFET) 77 equipment turnaround time  4 Escherichia coli  38, 254–256 ethylene vinyl acetate (EVA)  96, 98 ethylene vinyl alcohol (EVOH)  97, 98 European Directorate for the Quality of Medicines and HealthCare (EDQM) 162 European Medicines Agency (EMA)  136, 162 extractables  105–108, 143–144, 164, 215, 315, 318 The Extractables and Leachables Safety Information Exchange (ELSIE) 164

f Fate of Leachables  217 FDA’s Oncologic Drugs Advisory Committee 83 fed‐batch processes  282–284 field‐effect transistors (FET)  76–77 film manufacturing process biopharmaceutical manufacturing 101–102 blown film extrusion  100–101 cast film extrusion  101 extrusion lamination  101, 102 materials properties and type  99–100 monolayer film structure  99 multilayer film structure  99, 100 types 99

filterability studies  135 filter integrity tests  138–139 filter media  128–131 filter qualification and validation  139–140 filter selection applications orientation  136 flow decay studies  134–135 meeting process objectives  135–136 testing 134 filter train  134 filtration 123 FILTRODISC™ BIO SD  273 final sterile filtration  136–138 regulatory highlights  136 serial and redundant filtration  136–138 Flexel 3D Palltank  18 Flow‐through fractions  277 fluorescence spectroscopy  74–75 flux decline  135 Food and Drug Administration (FDA)  4, 83 foreseeable manufacturing scale  327 formulation and filling  6 challenges 287–288 end‐user requirements  288–289 hardware design and usability 290–292 quality by design  289–290 single‐use technology  292–295 freezing 18 Froude numbers  262, 267, 269 functional test–differential scanning calorimetry  109, 110 functional tests–assembly test  108–109 FUNDAMIX® SU system  32 fused deposition modeling (FDM) 240

g gamma sterilization  222–223 generic monoclonal antibody (mAb) process map  297–298, 300 glass fibers (GF)  128 good manufacturing practice (GMP)  39, 121, 197, 297, 299, 327 Green Chemistry Institute Pharmaceutical Roundtable (GCIPR) 175 Greenovation Biotech’s Bryotechnology 313–314

h Haemophilus influenzae fermentation  256–258 header biocontainer  294 heat deflection temperature (HDT) 244 heating ventilation air conditioning (HVAC) 303 helium integrity testing  115 helium test method  22 high‐capacity depth filtration  133 high cell‐density cryobag  190 high density polyethylene (HDPE)  95, 97 highly potent active pharmaceutical ingredient (HPAPI) safety and handling of  307–308 hollow fiber bioreactor systems (HFBSs) 38–39 host cell proteins (HCPs)  273, 274, 317 host‐specific, enzyme‐linked immunosorbent assay (HCP‐ELISA)  317 HxNy flu  83 hybrid production facilities  7 hydraulically driven mixing systems  27–28, 32–33 Hyoscyamus muticus 39 HyPerforma Single‐use Mixer (S.U.M.) 31

i ideally mixed system  26 IgG‐spiked CCS  277 immune cell therapeutics particular nature  329–330 technical standards required  331 techniques for cell expansion  331–332 uncertain mass production  330–331 imPULSE S.U.M., 31 infrared (IR) spectroscopy  74 in‐process controls  227 integrated continuous downstream processing (DSP)  124 integrity assurance  224–225 integrity tests  21, 112–113 intelligent control systems  87–88 intensified biomanufacturing facilities 184–187

337

338

Index

intermediate‐scale, single-use facilities 181 biomanufacturing facility  184–187 commercial manufacturing facilities 187 continuous processing  181–182 downstream processing  190–192 methodologies 183–184 moving toward intensified  181–182 process development  184 upstream process  187, 190 International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) 214 International Organization for Standardization (ISO)  301–302 invasive sensors  70 ion‐sensitive field‐effect transistor (ISFET) 77 isolator 290–291 ISPE’s Disposables Community of Practice 164

j JetMixer Bags  30 JM BioConnect  30

k Kymriah 83

l leachables  105–108, 143, 164, 215, 315, 317–319 leak test  6, 165 levels of change  228 Levitronix 57 LevMixer system  28–29 life cycle assessment (LCA) applied to SU technologies  173–175 early attempts  173 holistic methodology  172–173 LifeReactor 39 linear low density polyethylene (LLDPE)  95, 97–98 log reduction value (LRV)  134 low‐density polyethylene (LDPE)  95, 97 Lugaia 291

m MabXpure 277 Magnetic Mixer from Pall  29 mammalian cell culture  297–303 manufacturers’ integrity testing  224 master “wish” list  301 material‐dependent tests  105 maximum allowable leakage limit (MALL)  20, 21 meander‐type bioreactors  332 mechanically driven systems  27 mechanical tests  105, 106 medical grade  145 medium density polyethylene (MDPE)  95, 96 Meissner Filtration Products  31 melt temperature  233 membrane filters  131–132 mesenchymal stem/stromal cells (MSCs)  329, 331 microbial ingress  20 microporous filtration  127 depth filters  132–133 filter integrity testing  138–139 filter media  128–131 filter qualification and validation 139–140 filter selection  134–136 final sterile filtration  136–138 membrane filters  131–132 mycoplasma retentive filters  134 nominal vs. absolute removal ratings 128 particle retention mechanisms 128 sterilizing‐grade filters  133–134 virus retentive filters  134 microporous spargers vs. drilled hole spargers 252 MiniPerm 40 mixed cellulose esters (MCE)  128 Mixed4Sure 30 mixers 195 mixing 292. See also bag mixing systems definition 25–26 different impeller types effect  253–254 engineering principles  253 modeling and empirical measurements 253 quality 26 residence time distribution  27 Reynolds number  27

specific power input  27 time 26 Mobius CellReady  262–264 Mobius® CellReady 3l bioreactor  44 Mobius® Power MIX  30 modern operating systems  85 molding 102–104 vs. rapid prototyping  238–241 molecular fingerprint  74 monoclonal antibodies (mAbs)  305 mid‐scale production  8 monoclonal antibody drugs  83 moss cell line  314 moss manufacturing process  314 moss metabolite study  317 multitray cell culture systems  38–40 Multi Use Filling Machine  290 mycoplasma retentive filters  134

n National Technology Transfer and Advancement Act (NTTAA) 162 natural killer (NK) cells  329, 332 natural polymers  245 near infrared (NIR) spectroscopy  74 Nephelometric Turbidity Units (NTUs) 275 net present value (NPV)  211 noninvasive sensors  70 nonvolatile residue (NVR)  149, 151 normal flow filtration (NFF)  127 NovaSeptum system  64 N‐2 rocking motion bioreactor  190 nuclear magnetic resonance  151 nylon. See polyamide (PA)

o optical chemosensors carbon dioxide sensors  73 matrix‐embedded indicator  71 pH sensors  72–73 sensing oxygen  72 Optima MultiUse filling system  292 orbitally shaken bioreactors  46–47 aeration–interfacial area  265–267 flow dynamics  264–265 mixing dynamics  267 suspension dynamics  267–268 OrbShake® bioreactor  46 organic field‐effect transistors (OFETs) 77 oxygen enrichment of sparge gas  251–252

Index

p package seal quality tests  165 packaging and shipping validation  224–225 Pad‐Drive 31 PadMixer 31 Pall Biotech  292 Pall’s JetMixer system  30 Pall WandMixer  31 Parental Drug Association (PDA)  139, 159 particle image velocimetry (PIV) 264 particle retention mechanisms  128 payloads 305 PDA’s Technical Report No. 66  164 perfusion processes  284 peristaltic pump  55, 294 pharma grade  145 Physcomitrella patens 314 physical–chemical parameters  216 physical testing  105, 107 plastic polymers  231 plug‐flow reactors  123 PoGo G2 Mixing System  32 point‐of‐use integrity testing  225 point-of-use leak test (PoU‐LT)  22 polyamide (PA)  96–98 polyethersulfone (PES)  128, 132 polyethylene (PE)  95–96, 98 polyethylene terephthalate (PET)  97, 98 polypropylene (PP)  96 polytetrafluoroethylene (PTFE)  128, 131 polyvinylidene chloride (PVDC)  97, 98 polyvinylidene fluoride (PVDF)  128, 131 post‐use integrity testing  293 potentiometric sensors  76 power input  27 pre‐flush 147–148 pre‐use/post sterilization integrity test (PUPSIT)  138 primary disposables  5, 6 process analytical technology (PAT)  69, 84 process equipment‐related leachables (PERLs) 139 process model  297–303 product‐based tests  139–140 production bioreactor  190

Product Quality Research Institute (PQRI)  148, 159 protein pharmaceuticals  3 Pseudomonas fluorescens fermentation 256 PUPSIT. See pre‐use/post sterilization integrity test (PUPSIT)

q qualification  104–105, 228 quality by design (QbD) approach biological properties  231–233 materials, process and suppliers selection 231 process control strategy  235 specifications and process design space 233–235 quality insurance biological testing  105, 107 chemical compatibility tests  108 contamination requirements  110 endotoxin test  113 expiry date  110 extractables and leachables  105–108 functional test–differential scanning calorimetry  109, 110 functional tests–assembly test  108–109 incoming material control  112 integrity tests  112–113 material‐dependent tests  105 mechanical tests  105, 106 physical testing  105, 107 qualification tests  104–105 sterility tests  109 quality risk assessment  213–214 holistic approach to predict leachables 216–217 quantum 57 quarterly dose audits  223 QuattroMix Bags  33

r Raman spectroscopy  75 random co‐polymerization (RACO) 96 rapid prototyping  238–241 REACH 146 reduce–reuse–recycle 177 regulatory responsibility chart  146 residence time distribution  27

Reynolds number  27, 262, 269 Reynolds stress  262 rocking bag  269 room concepts  7

s Saccharomyces cerevisiae 255 SALTUS  31, 32 Saltus Vibromix Bioreactor  40 sampling for conventional systems  64–65 for single‐use systems  65–66 Sartorius Stedim Biotech  209 scanning electron microscope (SEM) 128 screw speed  233–234 sealing method  63 security of supply  6, 7 sensor patch  71 serial and redundant filtration  136–138 Serratia marcescens 38 single‐use bag  13, 18, 32 single‐use bioprocessing technology 313 single‐use bioreactors (SUBs)  6, 69, 84 conceptual design and software tools 237–238 engineering characterization, sterilization and qualification  241–245 features  37, 38 history 38–40 molding vs. rapid prototyping  238–241 orbitally shaken  46–47 plastic cultivation containers  37 selection 47 stirred 43–46 suitability 37 wave‐mixed bioreactors  40–43 single‐use continuous downstream processing capture and polishing  122 clarification 121–122 formulation 124 virus removal  122–124 single‐use devices  5–7, 331 single‐use diaphragm pump  57 single‐use equipment  147, 152–154, 195 advantages and limitations  4

339

340

Index

single‐use fermenters (SUFs) design criteria and approach  253–254 Escherichia coli fermentation  254–255 Haemophilus influenzae fermentation 256–258 heat transfer  249–251 heat transfer tests  254 liquid management  254 media sterilization  254 microbial 249 oxygen mass transfer tests  254 oxygen transfer  251–253 Pseudomonas fluorescens fermentation 256 yeast fermentation  255–256 single‐use production facilities  7–9 single‐use products  4 single‐use sensors  38, 92 single‐use standardization collaboration and alignment lead 162–163 compare and contrast  161–162 history 161 single‐use systems (SUSs)  4–7, 195 change management  227–228 component qualification  221–222 manufacturing and control  226 market (facilities) distribution  196–197 market estimates  197–198 market trends and perceptions  198–199 methods 196 operator training, performance culture 227 particulate risk mitigation  227 product design validation  222–226 single‐use technology (SUT)  159 alphabet soup  159–161 change notification  165 connectors 165 current industrial approach  214–216 design verification  165–166 downstream processing  120–121 evolution  169, 171–172 general efforts  163–164 holistic approach  216–217

LCA applied to  173–175 particulates 164–165 quality risk assessment  213–214 sustainability 169–171 system integrity  165 terminology and features  214 typical risks  214 user requirements  165 single‐use terminology  3–5 single‐use transfer lines connectors 55 fittings and accessories  54–55 pumps 55–57 tubes 53–54 valves and clamps  55 Smartainer II Shipper  18 smart sensors  86–87 Smartsheath 86 sol–gel technology  72 spectroscopic sensors fluorescence 74–75 infrared spectroscopy  74 interfaces 73 Raman spectroscopy  75 UV/VIS‐spectroscopy 73–74 SpinBag 30 stability studies  225–226 stainless steel (SS)  206, 209–211 facilities 197 stainless‐steel fermenters (SSFs) 249 Standardization of Single‐Use Components Extractables Studies 159 static mode  277 Steralloy FDG  245 Steralloy™ R‐resin  241 stereolithography (SLA)  241 sterility tests  109 sterilizing‐grade filters  133–134 stirred bioreactors  37, 43–46 flow dynamics  262 mixing dynamics  262–264 Reynolds and Froude numbers 262 scaling strategies  261–262 stirred tanks reaction  308 stock‐keeping units (SKUs)  178 sum parameters  149–151 supplier integrity test (SIT)  22 surface plasmon resonance (SPR)  77–78

surfactants 144 sustainability 169–171 implications 172 syringe pumps  57

t tangential flow filtration (TFF)  127, 198, 306 tank geometry  253 tank liners  13–14 TERUMO Quantum system  331 thawing 18 thermoplastic tubes  54 three‐dimensional (3D) bags  15 3D printing  241 total oxygen demand and transfer 251 toxicological concern (TTC)  151, 153 track‐etched membranes  128 traditional pharmaceutical products 3 TruBio DV  88 TruFluor 84 tube‐to‐tube fittings  54–55 tumor‐infiltrating T cells (TIL)  329, 332 turbidity 275 two‐dimensional (2D) bags  14–15

u ultra low density polyethylene (ULDPE) 98 Ultraviolet–visible spectrophotometry (UV/VIS‐ spectroscopy) 73–74 United States Pharmacopeia (USP)  110, 164, 165, 231 upstream process (USP)  315–317 intermediate‐scale single-use facilities  187, 190 tools 90 urethane casting  241

v ViroCyt® Analytics technology 206–208 virus retentive filters  134

w WandMixer 31 warhead 305 water for injection (WFI)  127

Index

Watson‐Marlow Fluid Technology Group 57 wave‐mixed bioreactors  40–43 WAVE Mixers  32 welding 59 Womersley number (Wo) 43

x

y

Xcellerex XDM Quad Mixing System 29–30 Xcellerex XDUO 2500 Mixer  30 Xcellerex XDUO Quad Intelligent Single‐Use Mixing System  30

yeast fermentation  255–256

z ZRP bioreactor  332

341