Process Analytical Technologies (PAT) in the pharmaceutical industry

Question 1: What are the main benefits of standardising PAT throughout the pharmaceutical manufacturing industries?

Don Arnone:

“A common set of standards to work to will both speed up deployment of PAT technologies and make the development and deployment of PAT tools more efficient for suppliers and the pharma industry itself. The setting of standards also helps manage expectations within both the supplier and pharma industries, and thus ensures that there is a good alignment of expectations for both sides.”

Dr Gawayne Mahboubian-Jones:

“The benefits are huge because we currently have a situation where product approval involves obtaining separate regulatory approval for every market you want to ship that product to. The cost of that is significant; in some cases where you have widely available products the cost is huge, not only of getting it up and running but the maintenance of all the regulatory approvals. The process of updating documents to accommodate newly discovered side effects also produces a major challenge; the level of correspondence required is immense and that creates huge overheads. Standardising the regulatory approach to PAT will reduce these administrative problems and it will become a lot easier to justify the use of PAT in the first place because the lower overheads for PAT processes will encourage companies to use them. Secondly, if you have a common approach then tech transfer from one site to another will become much easier, there are a significant number of products that the pharmaceutical companies want to move around the world for technical, marketing and commercial reasons. Companies with simpler processes are transferring them to sites where the cost of manufacture is cheaper, but the tech transfer is difficult because of varying regulatory approaches within different bodies. The benefits of common regulation are: dramatically reduced levels of unnecessary paperwork, reduced cost, easier tech transfer and an easier/faster response to issues that arise during manufacturing.”

Vinod Mehta:

“Well, this will take some time as PAT only started about 4 years ago. It is in its’ infancy and will take many years to mature.

In the implementation of PAT the benefits will be great, especially within process understanding and gaining insight into critical quality issues, as well as building quality into design and moving away from time consuming and costly off-line analysis. However, the shift can only be realised if there is parallel shift in moving away from unit batch production to a continuous process such as is currently in use for solid dosage manufacture.

In regards to the standardisation of technologies, it is unlikely to happen in the near future; for example in NIR there are many technologies that basically have a similar function but the core technology is different i.e. diode array, mems, FT, AOTF, scanning grating and filter. Standardised PAT software for all types of hardware will have a great impact on efficiency and operational cost.

Christian Woelbeling:

“I think it is too early to talk about a standardized approach.

A big benefit will be the change to a Design Space description including a control strategy, which is based on a deep process understanding using PAT technologies and ‘PAT Tools’. A further benefit of the PAT initiative is that the ASTM E55 standardizations and the Q-8, Q-9 and Q-10 Documents were pushed to a final state.”

Denise Root:

“The main benefit of standardising PAT methods is to ensure consistently high quality products from the pharmaceutical manufacturing facilities. The manufacturer will see increased productivity and yields of their products due to the real-time feedback control that is available using near-infrared (NIR) PAT methods. Standardised PAT methods will create a ‘common language’ to help communicate or transfer information between manufacturers. It could also help to ease the regulatory burdens, since approval would only need to occur once, rather than repeatedly. This one time approval would in-turn reduce the necessary paperwork and therefore speed up the entire process.”

Ingrid Maes:

“In order to achieve the full benefits of PAT systems, implementation of PAT tools require a great deal of indispensable interfaces with existing manufacturing and development architectures. At present, different automation and IT systems operate with different languages and cause both technical as organisational boundaries. Hence, interfacing standards do not only provoke an easier integration of different PAT tools & architectures, but also shorten the implementation/project time and allow quicker information exchange supporting Quality decision making.

Certainly from a PAT systems implementation project point of view, standardisation provides an environment that reduces implementation time (even for upgrades of different systems) and involved risks.

Standardising interfaces also has a benefit on business management decisions. By the ease of data exchange, a more efficient Quality data reporting, trend analysis and benchmarking of performance can be facilitated.”

Chris Hobbs:

“The use of PAT in the pharmaceutical industry will bring greater understanding as to why quality deviations occur both during the development and production stages of the product life cycle. This will lead to:

• More efficient processing – PAT supports a quality by design approach to manufacturing. PAT systems detect variations in critical quality attributes and adjust the process controls in response to these variances. This reduces the amount of variability in the process output, resulting in increased product yields, less scrap and rework, and making it possible to execute real time product release.
• A better quality product – greater process / product understanding allows for optimisation on quality. Higher product inspection rates and online inspection systems with feedback lead to better quality and more consistent products.
• Shorter time to market for products – analyser technology will give more visibility to the problems found when scaling up from small quantities produced in the lab to the large quantities manufactured in the production environment. Having the lab and production environment uses the same process analysis and PAT approach which makes it easier to transfer the knowledge learned during the product development phase to the production operation phase.

Standardising PAT throughout the pharmaceutical manufacturing industries will produce a standard body of knowledge that can be used by the companies and by regulatory agencies. This familiarity of analytical technology will result in more complete submissions, reduced review times and more efficient responses to any problems found with the process.

Price Waterhouse Cooper state that the use of PAT (i) Reduces Production Cycle Time by up to 40 %. (ii) Reduces production costs by up to 30%. (iii) Improves right first time prodcution from 90 to 99.9 %.”

Question 2: Which current PAT technology developments do you consider to be the most exciting?

Don Arnone:

“Infrared spectroscopy for content uniformity and related measurements has helped pave the way, but some of the newer technologies hold more potential. The ability of terahertz imaging to effectively provide knowledge of solid dosage form structure and chemical make-up below the table surface which is seen as key by Teraview and its customers to monitoring critical quality attributes of products both during scale up and in manufacture.”

Dr Gawayne Mahboubian-Jones:

“New instruments for use within PAT are constantly becoming available and these give us new insights into processes, but I wouldn’t identify an instrument as the thing I find most exciting; I see the way in which the application of PAT and companies’ assessment of their technologies and processes are changing as a real beacon of hope for the future of pharma industries. We are starting to see companies think very seriously about new technologies which are dependent upon PAT such as semi-continuous blending and drying. This approach is made possible by the way in which PAT gives people the flexibility to reflect on how we can improve and to be radical in the way we redesign processes; this redesign may result in the application of more sophisticated instruments, or in the demand for new measurement capabilities, but that is a short term factor. The exciting long term factor is that the industry is beginning to embrace change and improvement as a continuing part of production.”

Vinod Mehta:

“In my opinion the most exciting current PAT developments are; Diode Array, Mass Spec and Raman.”

Christian Woelbeling:

The FDA has to accompany every ‘PAT/QbD’ project during the submission phase. The new definition of process knowledge and process understanding leads to a completely different approach for the content of the submission documents and the QbD details shared with the FDA. Most of the PAT projects are projects in the existing manufacturing environments – and this is not the full PAT implementation starting with the submission phase. But these projects, dealing with PAT analytical methods, also provide a quick win by optimising pharmaceutical processes, though most of them were executed in pharmaceutical secondary manufacturing or fill and finishing. Due to the high value of products, a reduction of throughput times or reduction of stock lead to tremendous savings. This was experienced in the case studies evaluated by the ISPE regional PAT CoP initiative Germany/Austria/Switzerland when bringing 8 case studies for the ‘QbD/PAT Management Awareness’ Document together.

In general, this is the initiative started in the industry to get more process understanding. In fact, it is most exciting to see that the pharmaceutical industry is now thinking about how to improve processes and use tools to implement an optimised process control strategy based on QbD and PAT. There is no particular special technology, it is rather the fact that there is a lot of room for improvements and that the industry is moving.”

Denise Root:

“Near-infrared (NIR) technology is well suited for PAT. NIR is a rapid, non-destructive technique providing real-time results that can be used for process control and monitoring. Using NIR, exciting developments are being made in applications such as fermentation monitoring, fluid bed dryers/granulators and tablet analysis. With NIR, fermentations are being monitored/ controlled in real-time and in-situ, to optimise cell productivity. Novel sampling interfaces (i.e., NIR probes) are being designed specifically for use in the fluid bed dryers and granulators. Recently introduced NIR instrumentation, such as the FOSS MasterLabTM, allows the measurement of microgram levels active in tablets – levels that were unattainable until now. Watching these developments unfold is very exciting.”

Ingrid Maes:

“Several significant developments are already taking place on various aspects of PAT, and much more exciting developments are appearing in the near and long term:

• There is a need from the user point of view to have common PAT software, able to interface all PAT tools and manufacturing systems. Most of the current available PAT software is only covering a limited part of the PAT scope, often not integrated into one infrastructure. This makes it more difficult for use by non-specialists. At Siemens, we took a head-start on this and successfully developed and launched SIPATTM, which serves as the “glue” between various process analysers, Multivariate Data analysis, process control systems, LIMS systems, batch management systems and MES systems.
  Offering a common user interface for all these tools is one of the most exciting developments. It offers full flexibility to be used in process development but also in manufacturing. We have implemented SIPATTM successfully at a major pharmaceutical company where it is used in process development. The demand for PAT implementation at other companies is increasing exponentially.
• Another area where we see some changes happening as a result of PAT, is on the manufacturing strategy side. PAT enables the move from batch processing to continuous manufacturing operations and more and more companies are using PAT as a key component for a real-time product release system. Hence, a reduction of cycle time, faster process development and faster reaction to market demand are some of the many benefits which PAT can offer. Tighter integration with process control and IT systems is needed to achieve these benefits.
• We are convinced that PAT will be part of a Quality Management System, containing LIMS, SPC/MSPC (Model-Based Statistical Process Control) and including a real-time quality cockpit/dashboard. The latter is often part of an Intelligence Suite.
  We foresee a shift of load from LIMS towards PAT and SPC. Also we see integration with product specification management systems and product lifecycle management systems, starting from development and developing into manufacturing.
• We expect that Data Portals and Data Warehouses will become more and more important. For this, the standardisation initiatives will influence implementation and use, as alignment and interfacing various data sources will be required. On top of that, knowledge management systems will also be used in connection with PAT. FDA requests to demonstrate an increase in process understanding and knowledge can be supported with a KM system where tracks on knowledge progress can be kept.
• With regard to sensors and process analysers, more advanced analyser tools will be used, like process imaging, process microscopes, etc. However these systems generate enormous data volumes, urging the need for an IT system and data warehouse support. But also more advanced data mining tools are required.
• Along with the development of disposable process technologies (like a plastic bag as a replacement of a stainless steel vessel) there is a need for the development of disposable sensors and process analysers.
• Advanced analyser signals and MVDA (Multi-Variate Data Analysis) outcomes, are also requesting a more advanced process control strategy. APC (Advanced Process Control) knows a widespread application in many industries, though is rather new to the pharmaceutical industry. We see many initiatives and indicators that this will also affect the pharmaceutical industry.”

Chris Hobbs:

“For some time now the pharmaceutical industry has been aware of the benefits of applying PAT. The FDA effectively sanctioned the use of PAT when they published the “Pharmaceutical cGMPs for the 21st Century” in 2002. This guidance opened the way to a quality by design approach as opposed to a quality by test approach and it named PAT as a tool that should be implemented by the industry for this approach. In the five years since then, validated single process unit operations with PAT have been appearing in the industry, but the most exciting development is seeing companies move towards PAT platforms and tools capable of managing multiple analysers over multiple process units. These platforms store the data, both single point and array, in a regulatory friendly manner, securing the data and allowing data manipulations and calculations that can be used to make Good Manufacturing Practice (GMP) decisions. These developments are driving the need to create better standards for exchanging analyser data between all systems, opening the door to tighter integration between analytical devices and existing DCS/SCADA, Lab and Enterprise level systems. This integration is required to ensure that the benefits of PAT can be fully harvested. This will lead to better optimised Unit Operations with a better understanding of inter unit variations.

The ability of development teams, production organisations and regulatory groups to access and use data from every stage of a products lifecycle using Critical Quality Attributes (CQAs) rather than changing the process during scale up and having to find new correlations for the larger batches is very exciting. The use of PAT and its integration to existing systems will make this possible and will give full scale manufacturing, QA and regulatory groups a head start on preparing the process for commercial operation.”

Question 3: What improvements need to be made to existing products in order to further drive the use of PAT technologies in the market place?

Don Arnone:

“Ease of use and high throughput are key attributes for our customers. We also see the use of technologies in the design and development of new pharmaceutical products as key to their eventual use as PAT tools in manufacturing. More emphasis on deploying analytical technologies early on in the development of new pharma products is therefore required, and following these products through into manufacturing via PAT implementation of anlaytical tools. Applying analytical technology retrospectively into a manufacturing process for a pharma product which has reached the market is more difficult.”

Dr Gawayne Mahboubian-Jones:

“A lot of current PAT processes are based around laboratory instruments, we need to be able to ruggedise those laboratory instruments, not only physically but in terms of information, handling, range, scope, ability to withstand temperature variation, humidity change and vibration because manufacturing processes are real world processes.
We need to have an eye to the future about the time scales on which we expect instruments to respond, and the detection sensitivities. As we make processes more and more advanced the demands in these areas are going to increase.

The infrastructure around the instruments which binds them into the control of the process is also vital and development in this area by groups working together such as Optimal and Schneider Electric is equally important if we are to achieve the full potential instruments are making available.”

Vinod Mehta:

“Miniaturisation, lower cost of analysers and implementation/validation would make the technology widespread, and allow medium/small/genric pharma companies to make use of it. Common PAT software to interface to a number of different instruments offering one user interface for ease of implementation, training and use.”

Christian Woelbeling:

“We have to distinguish between using the PQLI/QbD/PAT initiative for new drugs on the one hand and for existing manufacturing processes on the other. The challenge is to calculate business cases upfront to identify the savings and than to change processes to apply PAT technologies. The management has to become aware that these changes towards more flexible PAT controlled processes are the key to efficient manufacturing processing.
The improvements for the industry are in using electronic systems.”

Denise Root:

“In order to further drive the use of PAT technologies, existing products must be more reliable, more automated, provide higher capacity and be better integrated into the manufacturer’s process control systems. Analyzers must be capable of running continuously 24 hours/day, 7 days/week, 365 days/year. The results from the analyzers must be integrated within the manufacturer’s control systems. Analyzers should be dedicated to the application with optimised sampling interfaces (as mentioned above) thereby enhancing the results for that application.”

Chris Hobbs:

“To date there are many ‘Islands of PAT’ single analyser solutions with their own unique interfaces. Because of this, there are as many interfaces and controllers as there are manufacturers of analysers. Some data standards exist but are not well established and are not used with great consistency. As a result of the lack of analyser interfaces and understanding of analytical data, DCS/SCADA manufacturers try to apply existing data management techniques to handle the analytical data, resulting in improper results.

In order to move forward, the industry must make it simple for users to be able to use multiple analysers from different manufacturers over multiple process units interchangeably. In order to do this there will need to be:

• A common standard to allow the extraction of single point and array data from analytical platforms.
• New requirements for DCS/SCADA systems to handle the different data structures produced by analytical devices.
• Analytical data and calculations will need to be stored in a regulatory compliant manner to ensure that they can be used in making regulatory decisions.
• Standard approaches to PAT validation must be created to prevent increased regulatory review times.
• A program for re-using PAT modules should be used to reduce the engineering and validation effort required to develop and implement PAT systems.”

Question 4: What future developments do you anticipate within the PAT industry?

Don Arnone:

“Additional moves towards standardisation of software, there is also evidence of additional endorsements of the PAT initiative by quality assurance and control functions within big pharma companies, which would provide an important driver for future adoption.”

Dr Gawayne Mahboubian-Jones:

“I think PAT will probably be accepted into the industry in 3 or 4 phases.

The first phrase will be the simple application of instruments to existing processes in order to learn how the process works; this will generate some significant improvements but it’s not going to take our manufacturing efficiencies up to the level of current the semi-conductor or petro-chemical industries.

The next step is to go back and redesign the processes, improving them by reducing the time that is often wasted so that they become leaner. At this level the infrastructure first starts to become really important, and having the correct expertise available to design that infrastructure is crucial.

Once the process is as lean as possible, the 3rd stage will be assessing the potential benefits of different types of process equipment; particularly if we look into the 21st century processes of the food and drink industries. We need to understand how these processes have changed and developed over the last 30 years; the concepts behind semi-continuous processes such as blending, granulation, drying and even brewing are available within these other industries and that does 2 things; first of all it provides us with higher quality products and secondly it dramatically reduces the size of the equipment we need. Instead of doing one batch every 4 hours with each batch taking 15 mins as you might do in large scale batch equipment, you are operating continuously over a 24 hr period but you are only ever working with a small amount of material at a time. This eliminates the need for a three storey industrial facility and allows for the use of a facility the size of a normal single person office.

In the long term, this has a dramatic impact in facility management, if your process has become that small, flexible and capable of relocating, then facilities in the next 20 to 25 years are likely to change dramatically. I think we will actually see facilities in the pharma industry which are much more like the facilities we see in the semi-conductor industry where, if your basic facility provides a clean environment you can move equipment in and out of that environment at will to make the process work the way you want. We may well reach the point where industrial processes are put together like building blocks, if one block doesn’t work another block can be inserted and when a block isn’t being used it can be out in normal storage, so the actual area you need to maintain at these very high levels of purity and cleanliness is much smaller; reducing costs of capital equipment and changing the whole cost basis of the industry.

Already the industry is looking at innovative ways of providing infrastructure, Schneider Electric are taking a lead in this where they can provide the entire operating infrastructure from a single source, liberating the pharmaceutical company to concentrate on its core expertise, and work is now ongoing with Optimal to ensure that PAT can be fully integrated into this capability. The developments in PAT will make the benefits of such single sourcing even more marked, but will require that the suppliers develop a much greater degree of PAT expertise to compete effectively.

Fundamentally we have a choice; unless Western Europe and the US actively pursue these developments aggressively over the next two years, other countries may well overtake us with the result that the pharma industry in Europe will disappear because it simply won’t be economical to produce here.

The conclusion is that the suppliers to the industry need to be just as innovative in their approaches to delivery as the pharma companies are in developing new processes.That requires a degree of partnership between the supplier and the industry which is only just beginning to develop (but which is much more advanced in other industries).”

Vinod Mehta:

“Development of standard methods for certain processes. Black box approach which requires minor tuning for process monitoring i.e. off shelf solutions which can go into the plant rapidly. The aim would be to have sensors such as temperature and pH but also provide chemical information.”

Christian Woelbeling:

“The implementation of PAT will bring a lot of innovation into the analytical industry. But the more important aspect for me is that the whole initiative is moving ahead, the new risk-based approach of the FDA for the 21st century started with the PAT Guidance launched in September 2004.

The dialogue with the industry began and some pharmaceutical companies followed the FDA initiative, starting several PAT projects. However, in the end it turned out that PAT was the wrong term for this ambitious paradigm changing initiative of the FDA. It moved over to a more umbrella-like view, to Quality by Design/QbD. But this also fails to cover the main goals of the FDA. In June 2007, during the ISPE Washington conference, the “Product Quality Lifecycle Implementation / PQLI” initiative was launched. This launch attracted not only 200 participants from the industry but also about 50 people from the FDA; during breakout sessions there were direct discussions between the FDA and the industry focusing on how to approach the new initiative and intensify the communication between the FDA and the industry. Dr. Moheb M. Nasr accompanied several breakout sessions; and in my view this was a full success for improving the dialogue between regulators and industry. The ISPE now has the opportunity to bring this further ahead. The basis for the PQLI initiative is set in the ASTM E55 and the ICH Q8, Q9 and Q10 documents. The benefits derived from this will be to have a process understanding that is coming along with a control strategy instead of only having a design space and a control space in place. Also, the 3-level model for control strategies, which was newly defined as a draft, will present a practical approach on how to handle patient- and business-critical aspects in submissions to the FDA.”

Denise Root:

“We anticipate development of higher capacity analyzers due to regulatory agency requirements of increased product testing (e.g., 100% raw material inspection). We also anticipate the integration of multiple analysis techniques. For example, FOSS has recently collaborated with Erweka, GmbH in the development of a commercially available automated tablet test station for measuring the physical tablet properties, as well as the chemical parameters (i.e., content uniformity and active level using NIR).”

Ingrid Maes:

“Besides other cited developments which have already been taking-off and which will be further deployed, we also expect big changes and developments in organisational and business aspects of PAT implementations. PAT requires multidisciplinary approaches and knowledge, but the maximum benefit of PAT can only be achieved when a broad PAT definition is applied. This will induce significant procedural and people change.

While FDA’s PAT guidance is not mandatory, putting these initiatives in gear can result in big savings for pharmaceutical companies. A Real-Time PAT system (including process instrumentation, data management & mining software and controls) interfaced with the manufacturing infrastructure allows real-time product release and prevents defective batches early in the process. Hence millions of dollars can be saved with relatively small investments. Companies need to be able to effectively communicate the strategic value of PAT investment to investors. They not only need to demonstrate the cost savings that will drive the return on investment (ROI), but also explain ways in which the company gains competitive advantage by applying PAT. PAT implementation strategies have to be outlined while demonstrating how PAT can be combined with other investment and improvement projects. The impact of different PAT implementation scenarios on inventory levels, work in progress (WIP), waste, quality, production time, lead time, people, etc. has to be demonstrated.

PAT is all about process understanding, being able to control critical to quality parameters and hence obtain a predictable process and pre-defined product quality output. Only when this condition is met can real-time product release and continuous manufacturing be achieved and consequently costs can be dramatically reduced.

PAT will enable the borders between R&D and production to fade. It is an interesting opportunity for the industry to be able to close the gap between development and manufacturing operations and technology transfer. Using PAT in development of quality surrogates for clinical performance (link critical product attributes to clinical outcomes) will be the next challenge to link clinical outcomes as part of the Quality System in manufacturing. This has the potential to accelerate the discovery and development processes with a more efficient and direct output. This gain in time (time-to-market) will lead to enormous financial gains.”

Chris Hobbs:

“Analytical devices will become smaller and cost less.

• Many more applications for existing analytical technologies will be developed.
• New analysers will be developed to measure process attributes that were never thought about before the PAT initiative was implemented.
• Improved analytical methods that will take advantage of increased analytical measurements and will be able to combine several individual analytical measurements into a single model that predicts a process output.
• PAT will be used in more processes, including more complex processes, as the industry discovers the advantages of full process understanding.
• There will be an OPC standard established for transferring analytical data between systems.
• Standard validated models will be developed to predict well understood quality attributes.
• Platforms and tools capable of managing multiple analysers over multiple process units storing the data, both single point and array, in a regulatory friendly manner will be the norm.
• Solution libraries will be developed to ease the deployment and validation time for PAT solutions.
• PAT data will define product quality.”

Manufacturing/ Academic panel

David Littlejohn, University of Stratchclyde

David Littlejohn received his BSc and PhD degrees in Chemistry from the University of Strathclyde. After a period of working at ICI, he returned to Strathclyde in 1981 and has been Professor of Analytical Chemistry at the University since 1988. He is currently Head of the Department of Pure and Applied Chemistry.

David has published over 140 papers and supervised 50 PhD students on various analytical chemistry topics including atomic spectrometry, chromatography, environmental analytical chemistry, conservation science and process analysis. His activities in the development and application of in-line, on-line and non-invasive methods of process monitoring cover a range of techniques and data analysis methods, including NMR, NIR, MIR, and Raman spectrometries, acoustic techniques and mass spectrometry. He was elected a Fellow of the Royal Society of Edinburgh (FRSE) in 1998, is a past winner of the Royal Society of Chemistry SAC Silver Medal, was awarded the Theophilus Redwood Lectureship by the RSC in 2001/2, and the Royal Society of Chemistry Award in Chemical Analysis and Instrumentation in 2006.

David is a founder member of the Centre for Process Analytics and Control Technology (CPACT), which has been in operation since 1997. CPACT is a multi-disciplinary industry-university collaboration devoted to research, technology translation and training in real-time monitoring, process optimisation and control. David has recently served on a number of RSE and RSC Committees, including the RSC Committee for Accreditation and Validation, and the RSE Research Fellowships Committee.

Pat Picariello, Director of Development Operations, ASTM International

Oversees the exploration, planning, organisation, & management of all aspects of new standards development activity for ASTM International. Picariello has experience with ASTM related to standards development and strategic standardisation initiatives both from a national and global perspective.

Picariello holds a B.A. from Dartmouth College and a J.D. from Temple University School of Law.
Picariello is a member of the Standards Engineering Society and the American Society of Association Executives.

Yatindra Joshi, Novartis

Yatindra M. Joshi, Ph.D., is Vice President of Pharmaceutical and Analytical Development at Novartis Institute for Biomedical Research, East Hanover, NJ. His responsibilities include Chemical and Pharmaceutical Profiling, Pharmaceutical Development, Life cycle management, Process Development/Scale-up and Validation, Analytical Development, Clinical Manufacturing, and Project Management functions. His interest/ expertise include dosage form design, application of statistical models for formulation & process optimisation, scale-up, and validation, drug delivery research, streamlining and accelerating drug development process, improving efficiency through robotics and automation, achieving organisational excellence, and building high performance organisation. Dr. Joshi received his BS and MS in Chemistry from Delhi University, Delhi, India, and Ph.D. in Chemistry from Dalhousie University, Halifax, NS, Canada, and MBA from the Wharton School of Business, University of Pennsylvania.

Staffan Folestad, Astrazeneca

Prof. Staffan Folestad, is a Senior Principal Scientist in AstraZeneca with responsibilities in the fields of Manufacturing Sciences, including PAT. He obtained his PhD. in Chemistry 1985 from the University of Göteborg. As an Associate Professor he directed research activities comprising development of Capillary Separation Techniques, Laser-Based Spectroscopy and Ultra-Trace Analysis. At 1994 he joined AstraZeneca (ex Astra Hässle). As a Senior Principal Scientist he has been leading global R&D activities in Process Analytical Technology (PAT) during a decade. As a Leader for the AZ Center of Excellence for PAT his research comprised development of advanced in-situ Process Sensors and in silico Process Modeling. Between 1999 and 2006, he held a part time position as Professor in Analytical Chemistry at Uppsala University. Among several external appointments he is member of the Royal Swedish Academy of Engineering Sciences, board member of the Center for Chemical Process Engineering at Chalmers University of Technology, and has been member of the EFPIA PAT Topic Group. He is also chairing the pan-European science conference EuPAT.

Steve Doherty, Senior Process Analytical Chemist, Eli Lilly and Co.

Steve Doherty is currently in the corporate PAT group at Eli Lilly and Co. Indianapolis. Steve received his B.A. in Chemistry at Wabash College in 1981 and his PhD. in Analytical Chemistry at Indiana University in 1989. Prior to joining Lilly, his industrial experience included four years performing field analysis with an environmental remediation firm, nine years of pilot plant and production plant process support at Dow Chemical, and five years in the process development group at Searle/Pharmacia/Pfizer. His current position involves the deployment of process analytical tools in development and manufacturing environments to increase process understanding, and partnering with instrument vendors to drive the implementation of new technologies. He is past Chairman for the Center for Process Analytical Chemistry’s Industrial Advisory Board, serves on the Scientific Board of the International Forum on Process Analytical Sciences, and is on the Steering Team of the Pharmaceutical Process Analytical Roundtable.

Question 1: What are the main benefits of standardising PAT throughout the pharmaceutical manufacturing industries?

Yatindra Joshi:

“PAT is an example of an area that is still being defined among the regulatory agencies throughout the world; the guidance is gradually becoming more clear, requirements for the ways in which reviews and inspections should be conducted are being implemented. If these requirements were standardised across the globe it would be easier for the PAT and quality by design principle industries.

From the industry perspective, standardisation also means the standardising of tools and their utilisation on applications. Standardisation would help in terms of the development of further tools and would make it easy for the industry to apply them to processes. A current problem within this area is that these standardisations are not widely applied, therefore many of the tools are not developed to the extent that we would like them to be; with the possible exception of NIR which is the most developed. Implementing standardisation would probably result in specific vendors putting more effort into the improvement of these tools because there would be more demand within the industry, or maybe multiple vendors would step up and improve these tools. Standardisation will provide regulation as have producing many other benefits across the industry.”

Pat Picariello:

“From the perspective of the standards developer, full consensus standardisation offers consistency, repeatability & reproducibility, increased user and consumer confidence, and lower production costs. The standards promulgated by ASTM International Committee E55 on Manufacture of Pharmaceutical Products are developed in an environment that is open and transparent. It (the process) offers the opportunity for all interested stakeholders to participate in an environment where all interests are equally weighted. In Committee E55, the global pharmaceutical community works together to develop a series of standards to help evolve the industry.”

David Littlejohn:

“I think that the need to standardise is not just across the large companies but across the small supplier companies as well; one of the problems we have is that if people are working to different standards and implement the FDA initiative to different degrees or perhaps for the small companies not at all, it will increase diversity in the ability of companies to fall on the guidelines that have been suggested. Something has to be done to try and ensure as much uniform uptake of this as possible. A particular problem with this is for the small companies because they don’t necessarily have the time/money or expertise to enter into the discussions and co-operations required in PAT in the way that big companies can.”

Staffan Folestad:

“A key AstraZeneca view and also my own personal view on PAT is that it is an enabler for a quality by design (QbD) driven development and manufacturing of medicines. Moreover, PAT is a key component when transforming an old pharma quality system into a modern drug quality system that is built on key concepts such as QbD and Risk Management. Standardising PAT is therefore more an acceptance of the paradigm shift for movement into a modern quality system; I think it’s also worth emphasising that when it comes to PAT tools and applications they always need to be product and process specific and will therefore likely be company specific. Still, the selection of tools will be scientifically based which means that standardising PAT as a ways of working benefits industry, health agencies and patients, whereas standardising PAT tools and applications should generally be avoided because there are several set of tools that are scientifically possible.”

Steven Doherty:

“To my mind, the benefits are two-fold. The most obvious is related to operational efficiencies. With the latitude to perform many assays in real-time, we should be able to achieve tighter control. Many of the metrics used in manufacturing, like asset utilisation, turnaround time, and labour costs, will compare favourably if we use the right PAT tools, and use them appropriately.

The other benefit is more philosophical, in that the PAT effort, along with GMP’s for the 21st century and quality by design, seems to signal a new way of doing things. The landscape has irrefutably changed in a major way for pharmaceutical companies, as well as for the agencies that monitor them. The market will not sustain the “old model” where we continue to do development and manufacturing as we have for decades. A fundamental shift is occurring, and the companies that can effectively integrate these changes into their business processes are much more likely to remain viable.”

Question 2: Is the ongoing support of the FDA proving beneficial in the practical implementation of PAT?

Yatindra Joshi:

“Absolutely, I think what the FDA did in August 2002 by introducing PAT guidance, encouraging communication between industry and the regulatory agencies and focusing on organisations like ICH, ISB and others has really helped. The FDA was also very active in the early days; creating forums and encouraging awareness to increase excitement. Over the years the FDA has changed, PAT has a limited meaning and is really more of a tool in the FDA’s current position, but how these tools are applied in order to improve quality by design is where the FDA has developed; and other regulatory agencies are following. What the FDA did is wonderful and where they are going will really help the industry to continuously innovate, improve and become more efficient.”

Pat Picariello:

“ASTM International activities are designed to mirror the respective industries they serve from a stakeholder participation perspective. The US Food and Drug Administration (FDA) is a key stakeholder in the global pharmaceutical community and as such, is extremely relevant to the work product developed by ASTM International Committee E55. The support and active participation of the FDA on Committee E55 has greatly contributed to the strength and quality of the E55 work product.”

David Littlejohn:

“My impression of talking to end user companies who are engaged with this process is that yes in general they do find that there is adequate support, there are certainly enough paper information available and also the various PAT meetings that different organisations arrange usually include at least one or two presentations from someone representing the FDA to explain or enhance paper information. My general impression is that manufacturers are reasonably happy that they can get information as and when they need it.”

Staffan Folestad:

“Well yes, the support of the FDA and other health agencies worldwide is key for establishing the new paradigm and introducing a modern quality system, it is also vital from a practical implementation point of view.”

Steven Doherty:

“The support of the FDA is necessary, but not sufficient. The agency has clearly articulated that they are offering us an opportunity to do things in new ways, but it fundamentally has to come down to practical business decisions. There needs to be clear business drivers and return on the investment for PAT to be relevant, particularly in a climate where resources are coming under additional scrutiny.

Perhaps as important as the support of the agencies is the mutual support of those of us in industry who want to see the success of this endeavor. To some degree, PAT has been used in pharmaceutical development and manufacturing for years, and it will continue to do so irrespective of the success or failure of the more formal PAT initiative. However, if we are to take full advantage of the opportunity presented to us, we need to help each other succeed. The “legs of the chair” important for long term viability are the industrial users, the agencies, and the instrument vendors, and we need to keep that perspective if we are to provide benefit to patients at the end of the day.”

Question 3: How is PAT used within your own laboratories?

Yatindra Joshi:

“PAT is used as a tool to gain better insight into processes and products, as well as for obtaining rapid information so that the processes and products can be monitored and controlled in order to achieve the desired qualities. On previous occasions these tools have not fully developed for utilisation in all aspects of processes that we currently have in the industry and our company. But there are areas in blending, tableting and drying where these tools are very nicely developed in terms of granulation. Improvement is still needed but these PAT tools are giving us better insights into our processes and products.”

David LittleJohn:

“Well of course we are a research laboratory interested in PAT as opposed to manufacturers who will use it for commercial reasons, so we either teach about process analysis and process control or undertake research into analytical methods or chemometric procedures etc, which are useful in process analysis and control.”

Staffan Folestad:

“PAT is embedded as a way of working and PAT tools are used in various types of applications from development through to manufacturing; for example, to characterise processes in order to achieve an in-depth process understanding. Another category of applications is for control which embraces these processes.”

Steven Doherty:

“Like most companies, we use PAT in development, tech transfer, and manufacturing control. Our principal focus is on development, because if we do a good job there, the benefits accrue through all the rest of the process. However, where there are clear safety, product quality, or business reasons, we have the resources to deliver into manufacturing as well.”

Question 4: Which current PAT technology developments do you consider to be the most exciting?

Yatindra Joshi:

“I think there are two; NIR is really very exciting and provides excellent information in terms of monitoring controlling. It’s been applied to multiple aspects, for instance with blending it fits very nicely with useful applications, it has also been very useful in drying, as well as in determining assay and continual formalities. We have now also demonstrated that you can use NIR tools for relating information to hardness and consequently dissolution. We get information through NIR that is much more reproducible than our conventional technologies, measuring hardness and dissolution which results in much more reliable information, we have only tested it for limited compounds and it therefore remains to be seen how well and how broadly it can be applied to multiple molecules, multiple products and processes. The other development is laser defraction which has proven very useful in terms of statistical aspects of design experiment and design multivariant/ analysis. These are not really true PAT tools but in my opinion they fall in the same area and have also been very helpful.”

David Littlejohn:

”Probably the move towards a new group of small companies new to the instrumentation market who are emphasising the opportunities of miniaturisation. I don’t think there is enough of that yet but if you are talking about what is likely to help reduce the cost of equipment and therefore encourage small companies to get involved in the use of technologies, it would be the production of lower cost/lower size high quality/performance instruments. Dome companies are now starting to look at this using developments in silicon based technologies etc, to try and take a fresh look at the design of PAT instruments, that’s probably the most exciting but it’s only starting and there’s a long way to go before it will have a definite impact.”

Staffan Folestad:

“Viewing this from an end to end perspective, for the customer, managing information in real time is key, on the other end driving new possibilities, emerging technologies and routes for development/manufacturing are the two areas which I personally find most exciting. As I am quite involved in academic research, new routes for advanced sensors that introduce new capabilities to monitor transformation of key material properties in-situ during processing is continually challenging. In addition, new tools for mechanistic modelling to enable simulation of multiphase process systems are very exciting. Clearly the industry faces challenges from the multi-phase process systems as used for formulated products as well as for bioprocesses; understanding these processes more fundamentally is on the whole the development of a more fundamental science platform and predictable modelling capability.”

Steven Doherty:

“A couple of the things that I find interesting are continuous operations and the shift towards micro-instrumentation. As with PAT, there have always been some continuous operations in pharma, but there are some types of chemistries that just aren’t feasible in the conventional batch mode that can be done in smaller-scale continuous mode. The extent of the published research in this area indicates that a lot of other groups are investigating the potential benefits of this approach.

Compact instrumentation and instrument systems is similarly a pretty active area of rapid development as well, and it seems that every trade show I go to there is a new compact analytical system. To some degree the developments in the system infrastructure (i.e. NeSSI, or the next generation sample system, a modular sample system footprint) are helping enable the sensor developments.”

Question 5: What future developments are necessary to ensure that PAT is utilised to its full potential?

Yatindra Joshi:

“The most important in my opinion would be regulatory acceptance and this again goes back to how we can standardise the usage of PAT across the globe in order to obtain one set of regulatory requirements and single standards for the way in which PAT is applied. The flexibility of the regulatory authorities will continue to improve, in my own presentations, the importance of this is emphasised; we have taken multiple steps in the discussion of PAT design space – part of a broader ‘quality by design’. The FDA is now talking about critical PAT initiatives, I think these are very welcome steps from the FDA; but if we want to apply these tools, the more flexible we make it for the industry the better. In my view, there should be more focus on the quality of the products as opposed to how we produce them. The progress would occur much more rapidly if the focus shifted in this manner, the process is not as important to a consumer as the end quality. If the regulatory industry becomes more flexible at some point in the future and more trust develops between the agencies and industry then the tools they utilise for processes would explode.

Another aspect of necessary PAT developments would be the standardisation of tools and vendors having a more active role in ensuring continuing development and improvement, as well as the industry taking the initiative to apply these tools to their processes and working with the regulatory industries to develop the application processes. Lastly, training and education of the people who would be involved in the development or application of these tools would be invaluable, as would training in the area of statistics and chemometrics.”

Pat Picariello:

“The continued commitment of the collective pharmaceutical industry (drug manufacturer, equipment manufacturer, user, regulator, and academician) will greatly assist in ensuring that the standards developed by ASTM International Committee E55 are as globally relevant as possible.”

David Littlejohn:

“There are two or three areas; the cost of online and inline instrumentation is really quite expensive, plus, the add on costs to actually install can be greater than the cost of the instrument by 2 or 3 times, so this is not an inexpensive action for a company to take, and that’s off-putting particularly to the organisations who have not got much experience in this area; they are not going to invest money in an area they are not sure about. So it would help a great deal if instruments were more accessible (in other words lower cost), the other thing that would help in relation to lower cost and size would be if instruments started to operate on a wireless mode; you would save on the miles and miles of cabling that has to be laid in an industrial plant in order to connect an analyser in a plant with a control room several hundred metres away from where that instrument is.

On the other hand, training is a huge issue, the big companies have got people who are able to spend time getting knowledgeable about instrumentation; what each type of instrument can do, what it can’t do, how its’ software develops, how to integrate data from different types of measurement and how to allow control. But the small companies (around 90% of chemicals manufacturers; the widest sense of chemical manufacturing in Europe are SME companies) do not have these resources. Without the opportunity to become knowledgeable they will not be able to buy this new equipment even if it’s cheap. So training is really very important and it has to be low cost, it has to be at the right level, it must be regional and it must be government sponsored in order to make an impact, otherwise there is going to be a huge increase in the gap between the relatively few big players in the game who are able to do it and the very large number of small players who will not have the confidence/time and money to get involved.”

“There is not a single answer to this question because it is a paradigm shift requiring parallel developments in several areas; to give some indication I would say that one key area is to achieve a true mind set in regulation, demonstrating commitment from Industry and Health Agencies to the new concepts. This is a challenge because it requires a learning exercise for everyone involved.

Staffan Folestad:

‘‘I think another area of necessary improvement is clear agreement in regards to appropriate content of PAT based submission dossiers, based on quality by design and risk management principles in order to move away from an old “rigid” assessment compliance culture into a more science and risk based regulation; which means that in practice it’s important to avoid developing a new checklist. The way in which people are working needs to be established with the use of a science and risk based approach, which would mean that a scientific point of view and assessment would inevitably be formed depending on the real risk dossier.”

Steven Doherty:

“I think we need to get to the point where the use of these tools and approaches is simply the norm, across pharma development and manufacturing. The two most important things I can think of to make this happen are an adequate supply of chemists and engineers who are conversant and comfortable with these tools, (which isn’t the case currently), and/or turn-key tools that enable non-practitioners to accomplish at least 80-90% of the target, so the “experts” can serve as consultants on the more vexing problems. There have been significant improvements in this area in the last several years, and some instrument vendors might take exception, but I don’t think the systems are quite there yet.”