Webinar Q&A: Qualification of PAT Based Control Strategies for Batch & Continuous Manufacturing
On- and in-line analysers are used in the framework of PAT (Process Analytical Technology) for controlling and monitoring unit operations in pharmaceutical operations for both batch and continuous manufacturing. The most basic requirement to utilize this technology in a GMP environment is that the instruments and their interfaces are commissioned and qualified before proceeding to the analytical method validation and productive use. A Process Analytical Technology system cannot be regarded as a stand-alone analytical lab instrument or as a piece of manufacturing equipment.
The purpose of this webinar was to provide practical examples highlighting different approaches how to commission, qualify and verify Process Analytical Technology systems designated for sophisticated control strategies in the most efficient and resource-saving way.
In the session below, our panelists respond to questions posed during the webinar.
Panelists:
Lorenz studied Chemistry and Mathematics and holds a Ph.D. in Physical Chemistry from ETH Zurich. He started his industrial career as an analytical scientist in chemical and pharmaceutical development at Roche and Novartis. Since more than 10 years he is working in the field of Process Analytical Technology and was involved in regulatory QbD filings. He is leading the Statistics and Process Analytical Technology group within global Manufacturing Science and Technology in Novartis Technical Operation supporting the production sites globally for all statistical relevant topics and managing Process Analytical Technology and QbD implementation with a strong focus on advanced control strategies based on NIR spectroscopy and multivariate statistical process control. Lorenz is the co-chair of the Process Analytical Technology & LCS Community of Practice.
He spent 7 years at Vertex, developing the Vertex continuous manufacturing (CM) platform, on which the first approved CM products are manufactured. He has over 25 years’ experience in the field having previously led the Global Process Analytical Technology Development Team, Pfizer Global Manufacturing. Martin is on the Exec Committee of ASTM E55, chairing ASTM E55.01 Process Analytical Technology, is on the ISPE Process Analytical Technology Community of Practice (CoP) global steering committee and chairs the UK Affiliate Process Analytical Technology Community of Practice. Part of this role at University of Strathclyde includes supporting the Centre for Process Analytics and Control Technology (CPACT) and CMAC Future Manufacturing Research Hub as well as being on Technical Advisory Committee (TAC) for the new Medicines Manufacturing Innovation Centre (MMIC) being built in Renfrewshire, Scotland.
Line is a physicist working as a Managing Consultant at NNE, an international engineering service provider with 1000+ professionals, specialized in pharma engineering. She holds a Ph.D. in NIR spectroscopy. Her background is pharmaceutical manufacturing and development. She works as a consultant focusing on Science & Risk based approaches in terms of QbD, Process Analytical Technology, Control Strategy, Process Validation, Qualification and Pharma 4.0. Line has been a member of ISPE for +15 years and has served many roles. She is currently the Chair of the global Process Analytical Technology & LCS Community of Practice
Q: How do you go about with the qualification for the vendor software of the Process Analytical Technology systems?
Lorenz: The software of the vendor is tested in terms of functionality as defined in the user requirements, e.g. 21 CFR Part 11. This is usually covered by the test protocols provided by the vendor. The software is expected to be validated and only the configuration and functionalities are being verified during IQ/OQ.
Q: Dr. Liesum mentioned method development was done on the qualified system. If this is done in parallel on a lab unit, what changes?
Lorenz: Qualification is a prerequisite for any development work foreseen for a method used in an GMP environment. This is relevant for both development and operations. If the NIR instrument is only used for development purposes, then no qualification is needed.
Q: Have you any advice on how to register Process Analytical Technology methods? Also, how to present/explain to Regulatory inspectors during audit?
Lorenz: Process Analytical Technology methods are handled as any analytical method with justification of specification and a summary of the validation and testing procedures reflected in 3.2.P.5. In particular for Process Analytical Technology methods development data and maintenance plans are expected as well (see also the relevant EMA and FDA guidelines)
Q: How to assure comparability method from Process Analytical Technology to pharmacopoeia testing (example: HPLC Method)?
Lorenz: Comparability is demonstrated by a method validation applying the same acceptance criteria as for an HPLC method, e.g. for precision. Accuracy is demonstrated by comparing the Process Analytical Technology with the HPLC methods. The acceptable bias depends on the inherent method variability of both methods.
Q: What about the integration of the Process Analytical Technology control system into the IT architecture
Lorenz: The prediction generated by the Process Analytical Technology system has to be transferred to other systems like Process control systems, global LIMS or archiving system. Therefore, an interface to these systems has to be established and verified during qualification.
Q: Is DQ necessary for off the shelf equipment?
Lorenz: In my opinion no DQ is necessary for COTS similar to benchtop analytical equipment.
Line: No, it is not, as you are not designing the equipment to meet your requirement. It is the other way around. You are justifying that a specific instrument or Process Analytical Technology system is applicable for your application and that is what you will verify.
Martin: It is quite normal to condense FS and DS for COTS/off-the shelf systems, and only DQ custom (non-COTS) functionality. Critical here is the validation master plan that gives traceability from requirement – functionality – design - qualification
Q: Could you please give an example of the influence of different Process Analytical Technology method risk loadings? i.e. what would change?
Martin: From a qualification point of view the low impact methods primarily focus on impact to process (impact risk assessment) and that they consistently achieve the desired performance. Medium/high impact methods have need also to demonstrate the desired performance in interaction with other systems.
Q: Have you any advice on how to register Process Analytical Technology methods? Also, how to present/explain to Regulatory inspectors during audit?
Martin: Engage early. For the US, request participation in the Emerging Technology Program (ETP) for platform decisions or Emerging Technology Team (ETT) for product specific advice. Have interactions and share intention before executing; audits of Process Analytical Technology systems are no different than any other aspects, expect questions around having systems/procedures in place and that those systems have been followed.
Q: How do you validate the script that need to be created for the Process Analytical Technology monitoring software to trigger data aggregation, analysed and predicted an outcome by a model, that would in send a signal back to the control system and trigger an actions?
Martin: If these are truly custom scripts then then they need to be generated/implemented under an approved quality system and introduced under change control including an appropriate level of script and functional testing.
Q: Should we qualify a PLC? Are IQ, OQ, and PQ required for a PLC?
Martin: A PLC is typically qualified under existing automation standards.
Q: If you are reliant on your supplier's factory testing, are you only shifting the responsibilities from internal testing to control of suppliers?
Martin: In part this is true, and it is important the vendors understand they become an auditable entity. However, think of these as a pyramid, where there are large efficiency savings by testing at source rather than every instance.
Line: The vendors/suppliers are experts in their technology/equipment and if you can trust your vendors there is no problem in relying on their testing programme and test execution. In pharma, we are re-testing way too many times without adding value to the equipment/process/business. The idea of a science & risk-based approach is to identify what needs to be tested and why (product dependent, control strategy specific) and make the test at the point in time where it adds most value. This assessment is based on the outcome of your risk assessment showing e.g. that moving the equipment will have no impact on the feature that will be tested. The test can therefore be performed before the equipment is moved and installed. In cases where the feature can be impacted by a move, you would do the testing later, but it can still be done by the vendor maybe as part of the SAT. This is a great example of an application of a science and risk-based approach. We leverage vendor documentation including testing as much as possible, if we can trust the vendor. To gain this trust, you may have to perform an audit of the vendors Quality System. It is still your responsibility that your process, equipment and control system is suitable for the intended use, but that does not mean that you must test everything yourself. Let the experts do the testing at the right point in time and location. You can find more information in the ISPE Baseline Guide Vol 5: Commissioning & Qualification 2nd Edition, and in the ASTM E2500 Standard Guide for Specification, Design, and Verification of Pharmaceutical and Biopharmaceutical Manufacturing Systems and Equipment.
Q: Should Cloud IOT integrated supply chain continuous production use the same base one-time validation or a continuous validation strategy?
Martin: In any qualification you are testing functionality that includes connectivity and integrity. Cloud IOT is no different; the difference is the connectivity, and so, integrity is at a different level or complexity and the use of verification on an ongoing basis matched is critical.
Q: The results from the poll were very interesting but how do we know how representative it is? Can you share the results and how many attendees answered? The results are less interesting if it was only answered by 10 persons.
ISPE: Poll results
ISPE would like to thank Line, Lorenz and Martin for their time to not only prepare and present their presentation but to answer these follow up questions. If you are interested in learning more about Process Analytical Technology systems, please join Line, Martin and others at the 2019 ISPE Annual Meeting & Expo on Tuesday, 29 October for the session titled Ensuring Pharma 3.0 before Moving to 4.0: Commissioning and Qualification of Complex Control Systems.
Line will lead this session which will present and discuss different approaches on how to commission and qualify (C&Q) manufacturing systems designated for Process Analytical Technology enabled control strategies including feedback control loops and a high automation level, e.g. using historians and MES. The discussion will be supported by case examples where an advanced control strategy was implemented, e.g. in the field of continuous manufacturing. In respect to Pharma 4.0, the impact of new approaches, such as artificial intelligence and self-optimizing systems, on qualification procedures will be elaborated.
