Platform Manufacturing of Biopharmaceuticals: Putting Accumulated Data and Experience to Work

Transcription

1 Platform Manufacturing of Biopharmaceuticals: Putting Accumulated Data and Experience to Work Table of Contents 1. Objective and Scope 2. Definitions 3. Introduction 4. Platform Manufacturing Approaches and their Application: Phases of Product Development and Commercial Product Manufacture 5. Leveraging Platform Process and Platform Manufacturing Data Throughout Drug Development Programs and Manufacturing Operations 5.1 Viral clearance / inactivation steps are usually platformed 5.2 A platform process generates data that are relevant to multiple products 5.3. How much data are required to specify a platform? 6. Presentation of Platform Process Data for Regulatory Review 7. Conclusions Contributors: Contributor / Coordinator: Contributor: Contributor: Contributor: Contributor: Contributor: Enda Moran, Pfizer Wolfgang Kuhne, F. Hoffmann - La Roche Ltd Kris Barnthouse, Janssen Biologics David Beattie, Merck Ranga Godavarti, Pfizer Piers Allin, EBE 1

2 Platform Manufacturing of Biopharmaceuticals: Putting Accumulated Data and Experience to Work 1. Objective and Scope Technology and process platforms are now used extensively in the biopharmaceutical industry, both throughout the various phases of product development and in the facilities used to manufacture clinical and commercial biopharmaceutical products. This document provides background information on the practical application of technology & manufacturing platforms within the biopharmaceutical industry. While examples and references are primarily made to proteins, the principles and approaches may equally be applied to other biopharmaceutical product classes such as conjugate molecular vaccines, live viral vaccines and so on. The value of data and knowledge gathered from platform manufacturing or platform-based process development is explored, and the advantages of the application and use of platforms, both to drug sponsors and the drug regulatory agencies, are considered. 2. Definitions Platform technology / platform process: a common or standard method, equipment, procedure or work practice that may be applied to the research, development or manufacture of different products Platform manufacturing: implementation of standard technologies, systems and work practices within manufacturing facilities, and their use for the manufacture of different products OR the approach of developing a production strategy for a new drug starting from manufacturing processes similar to those used by the same manufacturer to manufacture other drugs of the same type QbD: Quality by Design 2

3 IND: Investigational New Drug IMPD: Investigational Medicinal Product Dossier EBE: European Biopharmaceutical Enterprises Mab or mab: Monoclonal antibody HCP: Host cell protein CHO: Chinese hamster ovary GS-NS0: Glutamine synthetase-ns0 3. Introduction A platform technology or process may be generally defined as a common or standard method, equipment, procedure or work practice that may be applied across multiple products under development or manufacture. For example, platform may be used to refer to an expression system such as Chinese Hamster Ovary (CHO) cells, a high throughput screening system based on robotics, an analytical method such as capillary electrophoresis, a drug product formulation, a mode of cell culture such as perfusion culture, a process unit operation such as affinity chromatography or even a complete process comprising multiple unit operations, Figure 1. Figure 1 Examples of different types of platform-based tools or methods in biotechnology product development and manufacture. The term platform can be applied broadly. A cell line (expression system) Robotics Centrifuge A process Affinity Capture (ProA) Low ph Inactivation Ion Nanofiltration exchange (AEX) UF/DF 3

4 The use of platform technologies emerged in biopharmaceutical product development programs during the 1990s and primarily focused on monoclonal antibody production processes, Figure 2. As companies biologics portfolios expanded to more than just a few molecules, the benefits of consistency and similarity of tools and methods across product development programs was becoming apparent. There are multiple benefits of using platform approaches for development and manufacturing. Standardisation of approaches and tools across multiple products leads to improved quality and consistency, substantial cost-savings primarily as a result of more efficient resource utilisation (equipment/people), and faster process and product development. Platform approaches provide deep understanding of the impact of the process design and control stategy on process performance and product quality. This would have benefits in the implementation of QbD, and in understanding the likely impacts of manufacturing deviations or process changes. The use of platform technologies represents considerable opportunities to reduce risk to patient safety, providing access to platform technologies with proven performance and known safety profiles. Platform technologies applied to a particular product type, for example monoclonal antibodies (Mabs), introduces a degree of consistency to the products under development thereby benefitting their process and clinical development. Modern commercial biotech manufacturing facilities are now beginning to use platform technologies and standardised work practices to create efficiencies and flexibility, and to be capable of the manufacture of multiple products using a consistent set of well-established technologies. In the past few years, platform approaches have evolved to a mature state within manufacturing operations of the biopharmaceutical industry. For many reasons, principally economic ones, commercial manufacturing facilities of the future must be capable of the manufacture of multiple products. Some newer facilities constructed in the past few years will have this flexibility built-in, and will have the capability of manufacturing multiple products usually in a particular product family e.g., Mabs. The evolution of multi-product facilities will be enabled through the implementation of standard technologies and work practices within the facilities (platform manufacturing), 4

5 and the development of similar processes from product to product using standardised technology and unit operations (platform processes) for transfer into these facilities. Figure 2 The evolution of platform-based approaches in biotechnology product development and manufacture. In the 1980s, platform approaches were not employed. Currently, platforms are widespread within product development organisations and many production facilities. 1980s 1990s Companies often developing just one 1990s lead product Innovative and research-driven environment Little focus on manufacturability. Any of a multitude of technologies selected to make product No concept of Platform Biotechnology products approved New manufacturing facilities established Broad array of technologies used e.g., batch, fed-batch, perfusion (multiple systems), rollerbottle culture, different cell lines etc. whatever worked! Inflexible facilities First Platform approaches introduced within company development programs Mature Platforms within process and product development programs Many manufacturing facilities using platforms and standardised practices and capable of multiproduct manufacture New/different products (to a standard Mab) can still pose implementation issues for manufacture in established facilities Maturation of Platform Technologies/Processes 4. Platform Manufacturing Approaches and their Application: Phases of Product Development and Commercial Product Manufacture The pharmaceutical industry is continuously striving to increase the efficiency of process development in order to move promising products faster and more costeffectively into clinical studies. One preferred approach for rapid implementation of manufacturing processes for new biologics is to develop and implement early-stage platform processes that are designed to be suitable for the greater part of a company s product pipeline. A major intent of the approach is to prevent process development and clinical material manufacture becoming bottlenecks in the overall clinical development of new products. For example, production cell lines are created using the same parental cell 5

6 line, a similar vector and transfection method, and selected using a standardised approach (usually a high-throughput technology). The final choice of the production cell line would be based on its performance in the platform process using standardized culture media and culture conditions and the product would be purified using the same series of downstream process steps: this is different to past approaches to process development whereby a new process would be developed to accommodate the recombinant cell line. The product would be formulated in a standard formulation (fixed buffer composition) and presented in a consistent format, generally lyophilised or liquid. Process platforms should, by definition, meet a number of a company s pre-determined requirements regarding yield, process operability, product quality and so on for early stage development programs. A platform manufacturing process used in the clinical development phases would be expected to deliver the product requirements for supply of phase I clinical studies and would usually be satisfactory for phase II study supply. In some cases, a need for changes or improvements might emerge, e.g., to improve the robustness of the phase I process or to adjust or modify product quality attributes for phase II supply, but these changes would generally be slight modifications of the platform. Pivotal (phase III) clinical studies are usually supplied by the final version of the manufacturing process, and often at the intended manufacturing scale for commercial product supply. Late stage process development for phase III clinical material manufacture is usually based on the initial phase I/II platform process. Depending on the company s development strategy, the need for optimization, and the availability of alternative innovative technologies, single or multiple steps of the phase I/II process might be re-developed or optimised in order to specify the final, commercial manufacturing process. The platform process concept allows the reconciliation of what might, at first pass, appear to be incompatible requirements and outcomes of process and product development. That is, reduced costs and duration of development programs (a concern 6

7 for the drug sponsors, and the patients waiting for drugs) can be matched with increased process understanding, increased consistency of quality and heightened assurance of safety (a concern for patients, the drug sponsors and the Health Authorities). The quality of a new product can be assessed much earlier in the development process owing to the availability of standard platform analytical methods which would be evaluated, adapted and optimized on previous projects. In addition, the knowledge accumulated on previous projects may be leveraged in order to predict process capabilities under a given set of operating conditions. The increased process knowledge accumulated from several years of experience with a platform process therefore creates opportunities to employ QbDbased approaches to process development, enabling the definition of unit operation and process design spaces and at the same time, decreasing the development workload for any given project. These benefits are linked to, and dependent on, the implementation and use of effective data and knowledge management systems. Another opportunity is the concept of setting platform specifications for new molecules for those product quality attributes that are common across most, if not all, mab programs. The broad experience the industry has with multiple mab programs in terms of patient exposure means that the specification-setting exercise for a new program can be simplified for a number of standard attributes by defaulting to a template list of limits. This would primarily encompass those attributes that are related to safety of the product, for example purity, percent aggregate, percent residual DNA, percent residual HCP, and endotoxin (if normalized based on highest target dose). In keeping with regulatory expectations, these specifications could be relatively broad for phase I clinical programs, so that clinical trial data are acquired with product that has some range of attribute variability, and then systematically tightened as more manufacturing experience with the particular molecule is gained and process capability is understood. An extension of this concept would be to apply platform specifications even to those structural characteristics, such as oxidation, disulphide content, N-terminal and C-terminal 7

8 modification, glycosylation variants, etc., that have been shown through structure-activity studies to be not significant for the activity of a particular new product. Platform processes are currently most established in the biopharmaceutical industry for the manufacture of monoclonal antibodies. Monoclonal antibodies are regarded as well-characterized molecules and are known to exhibit very similar physico-chemical properties. In these cases, development and application of a platform process is regarded as cost-reducing, efficiency-increasing and in general, accelerative of programs. Most of the process platforms developed for monoclonal antibodies are based on a standard process structure: fed-batch cell culture usually in chemically-defined media, clarification by centrifugation and/or depth filtration, affinity chromatography capture, low ph viral inactivation step, followed by a combination of polishing steps including an anion exchange chromatography step (bind/elute or flow-through mode), a cation exchange chromatography step, a virus filtration step, and the final UF/DF and filtration steps. The main benefits of an established platform process such as this include; o Reduction of process development effort, time and costs. o Prior platform data informs risk assessments on new process weak-points, and focuses development effort where most needed. o Consistency in process performance and product quality (especially important when developing a particular class of products, e.g., monoclonal antibodies o Simplification of technology transfer activities to production facilities o Improved asset utilisation: one facility / same equipment for multiple products o Documentation preparation can be simplified; e.g., only minor adjustments to production batch records may be required from one process (product) to the next o Ability to translate learnings from one product to another as process database grows. Greater significance of a multi-product dataset as compared to a one-off study on a single product. o Raw materials and consumables are standardized allowing the use of materials with safety profiles proven to be acceptable, cost reductions through volume 8

9 purchasing and waste reduction as inventory stocks may be used across several different products o Reduction in time and resources leading up to and including pre-clinical and clinical studies o Reduction in failure rates during manufacturing due to accumulated process experience o Reduced personnel training burden owing to similar processes and equipment o Improved speed through repetition o Routine procedures for in-process and batch release testing lead to reduction of errors o Platform specifications for early-phase clinical programs o Broad database and experience to speed troubleshooting of manufacturing processes o Preparation of INDs/IMPDs/marketing authorisation applications may be facilitated more readily: pre-populated templates may be created which reduce the time necessary for manufacturers to author and prepare the submissions A platform process should not be regarded as a static, non-changing procedure. Its performance should be reviewed on a periodic basis. The manufacturer should ensure that the platform process is delivering a minimum acceptable level of process robustness, product yield, product quality etc. during clinical development of different products: these criteria would be developed based on a company s own specific needs. Changes to the platform should be introduced through controlled implementation of improvements and uptake of innovations. Such controls are important to understand the impact of improvements on the platform. 9

10 5. Leveraging Platform Process and Platform Manufacturing Data Throughout Drug Development Programs and Manufacturing Operations The biotechnology industry has matured to the point that many established drug developers and manufacturers can now claim considerable experience and knowledge gained through the development and manufacturing of multiple drug candidates. Extensive and valuable knowledge databases may be accumulated when platform processes and methods are applied to the development and manufacture of these biopharmaceutical drugs. Under these circumstances, the question is posed: can we leverage process and product data gathered across a portfolio of molecules, and apply these data to the development of similar molecules in the development pipeline? The benefits of such an approach have been discussed above and include speedier submissions if certain studies or validations do not need to be executed for each new product, faster file reviews at the regulatory agency owing to reduced data requirements in submissions, and greater significance and reliability can be attached to a multi-product dataset compared to a one-off study on a single product. Specific examples of platforms, the data derived from them, and the value of these data are discussed below. 5.1 Viral clearance / inactivation steps are usually platformed The viral clearance steps in a mammalian cell-derived biopharmaceutical manufacturing process are almost always considered when the use of platform process data/knowledge are discussed. There are good reasons for this. The same virus clearance studies are executed multiple times under the same protocol on different protein molecules. The different proteins, especially those from the same family of molecules, e.g., monoclonal antibodies, generally show similar behaviour at the different clearance steps (filtration/low ph or chemical inactivation/chromatography) which are operated the same way (platform process) for each product. These repeat studies tend to yield consistent and similar results from product to product. 10

11 Numerous published studies have demonstrated the capability of established downstream process steps for viral clearance. For example, the clearance of the simian virus type 40 (SV40) at an ion exchange step using Q-Sepharose Fast Flow (QSFF) resin and under different operating conditions was consistently > 5 log 10 for a number of monoclonal antibody preparations, Table 1. Table 1 Clearance of the SV40 virus at an ion exchange step using QSFF resin, under different operating conditions and for a number of monoclonal antibody preparations (mab 1, mab 2, mab 3). After Curtis, S. et al. (2003) Generic/Matrix Evaluation of SV40 Clearance by Anion Exchange Chromatography in Flow-Through Mode. Biotechnol. Bioeng., 84, 2, p179 Product Resin type Bed Height (cm) Linear flow rate (cm/h) Load capacity (g mab/l QSFF resin) Log 10 SV40 clearance ± 95% CI a mab ± 0.4 Naïve resin mab ± 0.4 Naïve resin mab ± 0.4 Reused resin b mab ± 0.0 Naïve resin mab ± 0.4 Naïve resin mab ± 0.4 Reused resin b mab ± 0.2 Naïve resin mab ± 0.1 Naïve resin mab ± 0.1 Reused resin b a Confidence interval. b The resin was reused more than 50 times under commercial operation conditions prior to being evaluated for SV40 clearance. The data in Table 1 are useful reference material to support the principle of generic or modular viral clearance steps, but a drug manufacturer must develop its own in-house data to support its case that a step that can be applied to multiple products without additional testing. The data in Table 2a are from a large drug manufacturer s development and marketed product portfolio, and show the inactivation of murine leukemia virus (MuLV) due to low-ph treatment of suspensions of sixteen different monoclonal antibodies. After experimentally executing low-ph inactivations for these 11

12 sixteen different IgG molecules, the step has been proven capable of achieving the minimum acceptable log reduction values (LRV) under a variety of conditions. These different conditions including buffer type, monoclonal antibody sub-type, protein concentration etc. were further examined in range-finding studies to assess the degree of robustness of the step, Table 2b. Taken together, the body of data is used to specify the conditions of operation of the platform step. This determination of the robustness in itself is a useful benefit to the manufacturer, as no further development or optimisation work needs to be executed on this platform step. However, there are opportunities to further exploit this rich database. For example, the data indicate there is little or no incremental value in executing a further inactivation study on another IgG protein in the pipeline: further data will only verify what is already proven. Therefore, there is an opportunity to eliminate testing and concentrate the saved time and expense on other aspects of the development process. Indeed, the recent CHMP Guideline on Virus Safety Evaluation of Biotechnological Investigational Medicinal Products (2008) acknowledges that prior experience and data may be used to support the reduction in virus clearance testing for investigational products under clinical development. For now, there are clear expectations for virus clearance testing identified in ICH Q5A (Virus Safety Evaluation of Biotechnology Products from Cell Lines of Human or Animal Origin), but this opportunity for streamlining and reducing unnecessary burden in the drug development and regulatory submissions process is an important one to consider for the near future. 12

13 Table 2a Low-pH inactivation of murine leukemia virus (MuLV) in suspensions of different monoclonal antibodies. This biotech manufacturer* developed a database covering a total of 16 different IgG molecules covering 2 sub-classes. Targeting an inactivation level of >4 LRV for this particular step. *Data supplied by member company of EBE Protein Class Number of different proteins tested Minimum LRV Maximum LRV IgG Table 2b Example of a bracketing or range-finding study of parameters that can impact the performance of the low-ph inactivation of MuLV. The biotech manufacturer investigated viral inactivation in experimental designs combining multiple of these factors and at the indicated levels of these factors. Data from these types of studies, and those shown in Table 2a, were used to specify a platform low-ph inactivation step. *Data supplied by member company of EBE Parameter / Factor Conditions mab Isotype 2 different IgG 1 2 different IgG 2 Protein Conc. (g/l) 5 20 Buffer type Acetate Glycine Salt Conc. (mm) ph 3.6 (worst-case of range ) Temperature ( C) 15 (worst-case* of range 15-25) 13

14 * Worst-case from perspective of inactivating an enveloped virus 5.2 A platform process generates data that are relevant to multiple products The data in Table 3 show the successful application of a platform method for cleaning of an ultrafiltration system post-use for the processing of two different monoclonal antibodies. In situations such as this where the protein concentration and buffer constituents are similar, the platform cleaning method negates the need to perform cleaning cycle development for each product, with concomitant savings in time, materials, documentation preparation, and automation system modifications. Table 3 Application of a platform method for the cleaning of an ultrafiltration system. Only two measures used to assess cleaning performance are shown. Measures are executed post-processing. Examples presented for two different monoclonal antibodies (Mabs). *Data supplied by member company of EBE Mab 1 Mab 2 14

15 Sample Type Acceptance Criteria Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 TOC Rinse 1.00 ppm Endotoxin Rinse 0.50 EU/ml <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 The data in Figure 3 illustrate the performance of a pilot-scale (thousands L-scale) platform purification process to deliver similar performance outcomes (in this case, the concentration of host cell protein (HCP) in the chromatography pools) across a panel of twelve different monoclonal antibodies. Operational confidence in a platform process such as this would be further assured through the use of multivariate experimentation in smaller-scale models of the critical unit operations to verify acceptable and robust performance when critical process parameters are varied. For many of the reasons detailed above, achieving this level of consistent performance in a platform process carries great advantages to the manufacturer and include the ability to plan for successful campaigns without concern for performance excursions or failures, the opportunity to speedily switch in-switch out products to the production facility and so on. The obvious question to pose is whether these data could be applied usefully to other products in any way, thus maximising the value of platform approaches and growing process databases. Figure 3 Levels of host cell protein in the chromatography step pools during the downstream processing of twelve different monoclonal antibodies. Abreviations: Pro A=Protein A affinity capture, BDS=bulk drug substance. *Data supplied by member company of EBE 15

16 For example (and referring to Figure 3), if a subsequent product (mab #13) is under manufacture, the opportunity presents to remove in-process sampling and testing at the ProA pool for the HCP attribute. Taken further, the process database may expand (more products) such that confidence of always meeting specification (or target) develops to the point that all sampling and testing for HCP could be removed. Indeed, and looking at it from a scientific perspective, a case could be built using platform and product-specific verification studies to remove the attribute from the final specification list of the substance or product. Note that HCP is used here as an example but the principle applies to any process or product-related attribute. Using platform process data in this way would represent a powerful application of data and risk management to the manufacture and testing of biopharmaceuticals. 5.3 How much data are required to specify a platform? There are a number of important and recurring questions to address when considering the robustness and generic applicability of data derived from platform operations or platform processes. Firstly, the drug manufacturer will at some point transition from developing the platform to considering the platform specified and 16

17 established: the question is when this transition occurs? The reality is that this transition will depend on the quality and amount of data the manufacturer has gathered on the platform step or process, and the type of process step or process under consideration. When we speak of quality of data, it is implied that certain types of data such as those from multi-factorial experimentation will provide greater insights into the degree of robustness and stability of a process or process step than those data from single factor studies. Regarding the amount of data question, and considering extreme scenarios just for illustration purposes, it would be difficult to claim a platform based on studies or process runs of just one or two molecules, but when the number of molecules studied expands to ten or greater as in the example shown in Figure 3 above, the platform could well be considered established. The amount of data required to verify a platform will also depend on the characteristics of a process step or unit operation. For example, an aggressive cleaning regime on an ultrafiltration system dirtied or challenged by a protein suspended in a simple buffer is unlikely to show substantially varied performance from one protein molecule to another. In this case, a performance evaluation across two or three molecules would provide sufficient evidence to the manufacturer that the platform cleaning step is robust and generically applicable to the processing of similar protein molecules. In contrast, much more process experience and data would be required if a platform fed-batch culture step was being developed for a particular class of molecule, owing to the biological (different cell lines, impact of certain raw materials etc.) and process variability inherent in this type of unit operation. In summary, there can be no rigid guiding rules on the amount of data and experience required to specify a platform process or process step. The drug manufacturer will always be in the best position to determine for its own particular cases what is appropriate and justifiable based on many considerations including prior process experience, degree of understanding of the factors (chemistry, biology etc.) influencing the performance of the process or process step, the ability to control the process or process step and so on. The second question concerns platform evolution, and asks whether a change to a platform step or platform process would constitute a minor modification to the existing 17

18 platform, or instead, signals the initiation of a new platform. Since there is a continuum of potential changes from minor to major, there can be no hard and fast rules as to what constitutes a transition to a new platform. For example, and at opposite ends of the spectrum, a few-degrees centigrade ( C) change in the temperature of the cleaning solution of an ultrafiltration system would very likely be considered a minor modification to a platform step, whereas a transition to a new mammalian cell expression system e.g., from CHO to GS-NS0, would very likely be considered a transition to a new platform process. Drug developers and manufacturers that have invested significant years and resource into developing process platforms tend to carefully modify and adjust those platforms in stages rather than make frequent, dramatic leaps into new platforms, since in the latter case, the lengthy phase of data gathering to assess process performance and robustness would have to start again. Finally, it is worth emphasising the potential value of the drug sponsor considering early communication with the regulatory agencies to discuss the justification and rationale for leveraging platform data, if there is to be significant reliance on platform process data in a forthcoming license variation or new submission. 6. Presentation of Platform Process Data for Regulatory Review Regulatory submissions efforts become more efficient through the use of platform process-based data and knowledge. Internal to companies, templates may be created for sections of IND/IMPD and commercial licensure submissions which reduce the time necessary for manufacturers to author the submissions. Regulatory agency reviewers will also appreciate submissions that make use of platform-based knowledge to demonstrate more complete understanding of the manufacturing process and the product. Site inspections, especially for multi-product plants, can also be simplified when platform approaches have been used, owing to the reduced number of unique manufacturing 18

19 processes and the concomitant reduction in complexity of the associated quality systems necessary to manage them. Platform data should also prove useful to streamline the filing process for postapproval changes to a manufacturing process or analytical method. For example, in instances where there is sufficient platform and product-specific evidence that changes to manufacturing steps and/or equipment or changes to analytical methods etc. have no impact on the quality, safety or efficacy of the product, then it may be possible to downgrade the classification of the post-approval submission. The application of platform-based experience and data to newer products under development without necessarily having to re-do every study for every product signals a maturation of the science, experience and knowledge underpinning biotechnological drug development. While there will always be some product-specific nuances to deal with in every development program, we are entering the phase of evolution of the industry where platform data are available to be presented for review to support clinical trial and marketing authorisation applications. These platform data packages might include those describing viral clearance capabilities of process steps, clearance of process related impurities, cleaning regimes for process equipment etc. How might these data be made available for review? Some possible means are considered here using a viral clearance step as an example (Figure 4), but as stated previously, the principles are not exclusive to viral clearance/inactivation unit operations. Current filing structures and mechanisms in place in many global regions at present do not work as described in Figure 4: this discussion below is intended to explore possibilities. Platform-process data could be compiled for multiple products to demonstrate generic applicability of the viral clearance step. Additional development data could also be presented, demonstrating the robustness of the step across the intended ranges of critical process parameters e.g., ph, temperature, time etc. and other process conditions providing assurance that the step performs acceptably within the operating ranges 19

20 intended for manufacture. This data package as a whole could be submitted for each product in each clinical trial application or alternatively, might be lodged in a reference document similar to a Drug Master File at the agency and would be available for consultation by reviewers at each new clinical trial application submission by the drug sponsor. The reference document or Master File could be updated by the drug sponsor if a significant process change was introduced to the viral clearance step, or if a new replacement clearance step was introduced. If available data were insufficient to support the generic applicability of the changed or replacement clearance step across multiple products, product-specific virus clearance data would be submitted under normal practice until data accumulate to support the step as a platform or generic step. The Master File would then be updated with new data, and the package applied to subsequent products under clinical trial or marketing authorisation application. 20

21 Figure 4 Possibilities for presentation of platform-derived data for regulatory review. A generic virus clearance data package is shown as an example, but this illustration could apply to any package of platform process data. Data could be presented for review by submitting the same data package with each new clinical trials application (Path A) or presented once to the agency to be maintained as a type of Master File (Path B). The principles shown could equally be applied to Marketing Authorisation Applications. 7. Conclusions The evolution and benefits of platform technology and processes throughout biopharmaceutical drug development and manufacture have been discussed. The benefits of using a platform approach are apparent: standardisation of approaches and tools across multiple products leads to improved quality and consistency, substantial cost-savings, efficient resource utilisation (equipment/people) and faster process and 21

22 product development. Modern commercial manufacturing facilities use standard/platform technologies and work practices to improve efficiencies and flexibility, and to be more capable of the manufacture of multiple products. Platform processderived data at many drug developers/manufacturers have accumulated to extensive databases. The biopharmaceutical industry, including the regulatory bodies/health authorities, is now in a position to exploit these data to improve biopharmaceutical drug development, manufacture and regulation. 22