Documenting the Journey: Assembling a PAT Dossier

As Albert Einstein once said, “If you are out to describe the truth, leave elegance to the tailor.” A PAT dossier, describing your process from a scientific and technological standpoint, will be the log for your technical journey. This article summarizes what should go into this important record.

By John R. Davis, PE, president, and John Wasynczuk, Ph.D., senior vice president, Lloyd’s Register Serentec, Inc.

What is a PAT dossier and why do you need one? Before we answer that question, let’s begin by examining the tasks and activities that need to be completed before you are ready to implement PAT-based quality manufacturing.

When you are implementing Process Analytical Technology (PAT), one of the end results is the collection and approval of all the assumptions, rationale, experiments, statistical analysis and conclusions required to demonstrate to an outside agency (such as FDA) that you have a thorough process understanding that enables you to implement a PAT application. We’ve all heard it said, “Once you know your destination, you can plan your itinerary.” As with any type of undertaking, it helps if you start with the end in mind. So let’s start at the endpoint and work backward.

Imagine that you have just completed a PAT project. Your new process is so well characterized, monitored and controlled that you no longer require end-product testing. All quality attributes are assured by careful monitoring using sophisticated in-line, on-line and at-line analyzers and tightly controlling all parameters using state-of-the-art model-based multivariate control.

With each successive run, you add to your overall knowledge of your process and the data collected is cycled back into your model so that the entire system is continuously improving. Quite an achievement! So, what did it take to get to this point? You had to start with small steps. There were many hurdles, many decisions, many experiments, and exhaustive review of results. Some experiments failed and some succeeded, but you learned something from each of them.

Analyze processes: Your PAT project undoubtedly began by looking at the existing process and analyzing existing data. This helped you identify opportunities for improvement and process steps that were poorly characterized or troublesome.

This data was collected from many sources (lab, maintenance records, investigations, etc.), organized, fed into a database and evaluated. From this data, several process steps were selected as candidates for PAT. This exercise also gave you insight as to what parts of your process were not well understood and needed more investigation. A review of this data and the decisions made needed to be documented, and so began the PAT dossier.

Define critical quality attributes and identify critical control points: Before going any further, you decided that it was important to identify and document what exactly were the important outcomes of this process and what were your current critical control points. So you consulted with everyone associated with the process and put together a list of all important quality attributes and control points. These were then categorized as to level of importance and risk. When all stakeholders agreed, you decided that this was an important accomplishment so you placed the results in your PAT dossier.

Design of experiments: As you reviewed your process, you began to identify gaps in your overall knowledge of the process which led you to begin formulating experiments that would help fill in these gaps. These experiments had to be carefully designed to minimize the number of experiments needed and maximize the information gathered from each.

You began with some screening experiments to determine which factors were most influential and that they were within their appropriate ranges. You then set up some optimization experiments to determine if your process had a unique optimum for all quality attributes or if it would be necessary to find a compromise among conflicting quality attributes.

Finally, you ran some robustness experiments to determine how narrow your tolerances would need to be to maintain an acceptable level of variance. A lot of thought and preparation went into the design, execution and analysis of these experiments. You felt that it was important to communicate the rationale and logic used for these experiments, so you decided that the PAT dossier was the appropriate place to document all of this information.

Develop statistical models: The amount of data collected from the experiments was overwhelming. You needed a way to interpret and visualize the data to better understand the secrets they held, so you enlisted the help of software. This software generated many different charts and representations of the data that provided the insight into parameter relationships that you were seeking.

Several of these representations clearly depicted the signature of your process and were so helpful in increasing your understanding that you thought they would be equally helpful for others, so you printed them out and added them to your PAT dossier. Of course, you were aware that this base of information would increase over time and you would need to revisit the data and its representations periodically, but you felt it was important to document your original findings.

Research analytical technology solutions: Armed with this new understanding of your process, you set off to research the available technology that would assist you in monitoring and controlling key parameters. You investigated several possible options and put together a table to compare characteristics of each. You looked at things such as reliability, accuracy, availability, maintenance, response time, and costs. You visited locations where the technology was being used and asked many questions to satisfy yourself that you were making the correct decisions. Once again you thought it would be important for others to understand the basis of your decisions, so you put all this information into the PAT dossier.

Contents of the PAT dossier: As time went on, you found that there were many scientific and technical decisions, conclusions, rationale and other important items to document, including:

  • The extent of experimentation
  • The interpretation of results
  • The accuracy of mathematical models
  • The capability and reliability of the newly implemented technology
  • The assessments of risks
  • The selection of critical control points and critical quality attributes
  • The selection and placement of analyzers
  • The selection and configuration of software
  • The determination of endpoints and process ranges (i.e. demonstration batches
  • The results of demonstration batches
  • The determination of process robustness
  • Confirmation of variance reduction
You felt strongly that since you were running your process based on a scientific understanding of its character, you needed to have a compendium that captured all of the science and technology that went into it. Hence, you created your PAT dossier. Now you find that this compendium is invaluable. You have used it repeatedly for many purposes, such as:

  • Convincing management that you were ready to implement PAT
  • Convincing QA that the new system would provide much better control
  • Defending your PAT-based control to FDA and other auditors
  • Training managers, supervisors, QA, operators and maintenance
  • Training your replacement when you were given that big promotion
Albert Einstein once said, “If you are out to describe the truth, leave elegance to the tailor.” The central purpose for assembling a PAT dossier is to describe your process from a scientific and technological standpoint. This document needn’t be polished and shouldn’t be verbose, but it should be able to stand up to scientific scrutiny. There should be more emphasis on accuracy than format, on content than length, on insight than impression. As this industry travels toward a more scientifically-based destination, documents such as the PAT dossier will become the log of your technical journey.


References

1. FDA, “Guidance for Industry: PAT – A Framework for Innovative Pharmaceutical Manufacturing and Quality Assurance. Draft Guidance.”

2. Jones, Bradley and Creighton, Lee, “Data Management & Analysis”, R&D, September 2004.

3. Eriksson, Johansson, Kettaneh-Wold, Wikstrom, and Wold, Design of Experiments, Stockholm, Sweden, Learnways AB, 2000, pp 3, 10



About the Authors

John R. Davis, PE, is president of Lloyd’s Register Serentec and has more than 20 years of experience in the design, commissioning, startup, and validation of pharmaceutical and biotechnology facilities. Prior to founding Serentec in 1996, he held progressive positions with Glaxo Wellcome Inc., Burroughs-Wellcome, Fluor Daniel and Pitman Moore. Davis has been involved in process automation and controls systems design, implementation and validation in the biotech and pharmaceutical industries since 1982. Earning a B.S. degree in Chemistry from Indiana State University, Davis went on to become a registered professional engineer in Indiana and North Carolina. Professional organization memberships include ISPE, PDA, ISA and ASQ.

John Wasynczuk, Ph.D., is senior vice president of technical services for Lloyd’s Register Serentec and has more than 22 years of manufacturing operations, validation, regulatory compliance, preparation of New Drug Applications and FDA Pre-Approval Inspections. Prior to joining Serentec, he held validation and production operations positions in sterile production facilities at Novo Nordisk, Burroughs-Wellcome and Pfizer. Dr. Wasynczuk has been involved in several new plant start-ups and has experience in formulation, filling and packaging of sterile powders, solutions, suspensions and ointments. He is a member of PDA and ISPE and has a Ph.D. in Microbiology from Miami University.

Show Comments
Hide Comments

Join the discussion

We welcome your thoughtful comments.
All comments will display your user name.

Want to participate in the discussion?

Register for free

Log in for complete access.

Comments

No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments