Guidance for Industry Process Validation: General Principles and Practices Dr. Mark Tucker, F. Hoffman-La Roche, Ltd.

Disclosures I am currently a Senior Technical Advisor at F. Hoffman -La Roche. I worked at the U.S. Food and Drug Administration (FDA) from 1996 - 2002. My last position at FDA was Director, Investigations Branch, in the Los Angeles District. The following are my views and not necessarily the views of the Food and Drug Administration Alumni Association (FDAAA), FDA, or Roche. Expenses for travel are being paid by Roche. FDAAA permits the reuse of these slides for educational purposes with attribution to the creator and FDAAA.

Presentation Overview Guidance Background GMPs and Process Validation Process Validation Stages – the Lifecycle Approach

Background Issued January 25, 2011 Furthers the goals of GMPs for the 21st Century initiative by fostering innovation and advancing science in pharmaceutical manufacturing. Aligns Process Validation activities with the product lifecycle More rational, scientific and can help improve control and assurance of quality Approach aligns with Quality by Design (QbD)

cGMP Regulations for Finished Pharmaceuticals Process Validation is an enforceable requirement for finished drug products: 21 CFR 211.100(a) “written procedures for production and process control designed to assure that the drug products have the identity, strength, quality, and purity they purport or are represented to possess. 21 CFR 211.110(a) “… procedures shall be established to monitor the output and to validate the performance of those manufacturing processes that may be responsible for causing variability… Explain that these two key regulations are our basis for requiring PV. Point here is that a careful examination of these two regs support the need for proper DESIGN of the manufacturing process and that companies need to be knowledgeable of sources of variability in the process and monitor the process so the right product meeting its specs and quality attributes is always produced. Okay, so that clears up finished product... for APIs... [next slide]

APIs and the FD&C Act Explain that there are no regs for API. FDA recognizes ICH Q7 as the CGMPs for APIs. [On APIs - Just a note The validation section in Q7 does not sync up with this new draft Guidance so be aware of that. I wouldn’t mention it but acknowledge it if someone in your audience does.]

Process Monitoring 211.110(b) Valid in-process specifications for such characteristics shall be consistent with drug product final specifications and shall be derived from previous acceptable process average and process variability estimates where possible and determined by the application of suitable statistical procedures where appropriate. Examination and testing of samples shall assure that the drug product and in-process material conform to specification. In part, established the need to monitor and analyze process

Statistical Analyses 1978 Preamble1 response to comments regarding 211.110(b): “Further, after product histories are developed, the Commissioner encourages manufacturers to perform statistical analyses on their products and processes with a view to controlling batch-to-batch variability to the maximum extent possible.” 11978 Preamble, Human and Veterinary Drugs, Good Manufacturing Practice in Manufacture, Processing, Packing, or Holding As a regulator, this is easily my all time favorite quote from the preamble to the 78 revisions [EXPLAIN PREAMBLE PURPOSE]

Sampling 21 CFR 211.165(d): Samples must represent the batch being analyzed. (21 CFR 211 160(b)(3)) Meet specifications and appropriate statistical quality control criteria as a condition for batch approval and release. (21 CFR 211 165(d)) Another GMP reg with statistical reference. Don’t’ get too heavy into this one because it’s really talking about batch release. We’re talking about validation of a repeated process.

More on 211.165…. Section 211.165 is therefore modified to allow greater latitude in establishing acceptance criteria, while retaining the basic requirements that acceptance criteria for sampling and testing, and for acceptance levels, be based on appropriate statistical quality control criteria. The Commissioner is convinced that sound statistical methodology should be applied to the procedures for testing of attributes or variables that impact on the quality of drug products and the evaluation of the results of such testing to determine acceptance or rejection of the lot. The uses of AQL and UQL are examples of statistically derived levels for acceptance or rejection. The Commissioner believes that more study must be given to this aspect of manufacturing practice and advises that in the future FDA will invite additional industry comment regarding revision of this section This slide contains text from the preamble on 211.165(a). Read the highlighted part. Just reinforce that the concept that for many years since the last major GMP revision, the agency has held that sound statistical methodology should be used and we may have let that fall through the cracks in some cases with our PV endeavors.

Data Analysis Requirements Section 211.180(e) requires that information and data about product quality and manufacturing history be periodically (at least annually) evaluated to determine the need for changes in specifications or manufacturing or control procedures, and must include: A review of a representative number of batches, whether approved or rejected, and, where applicable, records associated with the batch. A review of complaints, recalls, returned or salvaged drug products, and investigations conducted. Ongoing feedback about product quality, and a process to react to that feedback, is an essential feature of process maintenance. Process validation documents are associated with every batch. Continues theme that processes should be monitored and data analyzed as part of continual improvement.

The Questions of Process Validation What scientific evidence assures me that my process is capable of consistently delivering quality product? How do I demonstrate that my process works as intended? How do I know my process remains in control ANY QUESTIONS THUS FAR ESPECIALLY AS CONCERNS THE REGULATIONS SUPPORTING THE NEED TO VALIDATE MANUFACTURING OPERATIONS? Now on to an overview of the recently revised guidance....

The ‘process’ of Process Validation Process Validation is defined as the collection and evaluation of data, from the process design stage through commercial production, which establishes scientific evidence that a process is capable of consistently delivering quality product. It is a series of activities taking place over the lifecycle of the product/process. Not a one time event but key milestones. PV is ongoing but there are key milestones, the biggest one being that point at which a company decides their initial process design work and process qualification (as referred to in the new Guidance) work is sound and solid enough that they have reached “a high level of assurance that…..see PV def in new Guidance.

Lifecycle Approach Lifecycle Overall validation is not “completed” but ongoing Necessitates comprehensive process design to understand sources of variability and achieve process understanding Incorporates risk management Recognizes that more knowledge will be gained during commercialization Main point is that we encourage Risk Management in all PV activities. Did not elaborate in the Guidance because it is fully discussed in other guidances out there such as ICH Q9. We know you will gain more process knowledge once in routine production and that info may feed back into process design if necessary.

Process Lifecycle Stages Stage 1, Process Design: Lab, pilot, small scale and commercial scale studies to establish process Stage 2, Process Qualification (PQ): Facility, utilities and equipment qualified Process Performance Qualification (PPQ) Confirms commercial process design at (or near) scale Stage 3, Continued Process Verification: Monitor and assess process during commercialization for: process improvement assurance that process remains in a state of control. Can just read this. We picked 3 stages. Can be overlap. Emphasize that Stage 2 loosely correlates with our old habit or practice of PV but emphasizes more PERFORMANCE criteria in addition to batch sampling and testing.

Goals Stage 1 - Functional understanding between parameters (material and process) and quality attributes Stage 2 – Measurable scientific evidence that product will consistently meet specifications process performance meets acceptance criteria; reproducible Stage 3 - Maintain or improve control and reduction in product and process variability While each Stage has it’s important aspects, the main goals of each stage are summarized here.

Stage 1: Process Design Propose process steps (unit operations) and operating parameters to be studied. Identify sources of variability each unit operation is likely to encounter. Consider possible range of variability for each input into the operation. Evaluate process steps and operating parameters for potential criticality. Activities that are typical of Stage 1. Occurs concurrently with product development.

Stage 1: Process Design “Focusing exclusively on qualification efforts without also understanding the manufacturing process and associated variations may not lead to an adequate assurance of quality.” The poor quality drugs on the market as evidenced by the number of complaints, recalls and other indicators, from supposedly validated processes, indicated a a lack of process understanding, process control, or both, and was the impetus for revising the ‘87 guideleine.

Stage 1: Process Design Manufacturers should understand the sources of variation, measure the degree of variation, understand its impact on the process and product quality attributes, and manage the variability in a manner commensurate with risk it represents to the process and product. Develop mechanisms for managing variability is part of the control strategy really about describing a relevant and appropriately protective control strategy designed to ensure uniform material meeting target specifications can be consistently mfd with detection/mitigation measures should something fail or go not quite right, may choose advanced manufacturing technologies that employ detection, analysis and control feedback loops to react to input variability.

Process Design Outputs Master production and control records Overall control strategy Operational limits/ranges Specifications What comes out of your design efforts? These items.

Stage 2: Process Qualification (PQ) Process Qualification: provides confirmation that the process design is functional for commercial scale manufacturing. Transfer process design knowledge to production, i.e., technology transfer This is Stage 2 and here the company is confirming or verifying that the full scale process will perform as anticipated, designed, predicted.

Process Qualification Two Aspects Qualification of equipment, utilities and facilities Process Performance Qualification (PPQ) This is Stage 2 and here the company is confirming or verifying that the full scale process will perform as anticipated, designed, predicted.

Facilities, Utilities and Equipment Precedes PPQ Consider user requirements, use risk analysis to identify studies/ tests needed and chose criteria to assess outcomes Plan for handling changes Generally engineering with development, production, and quality unit involvement Quality Unit reviews/approves the qualification plan(s) and report(s) Activities performed to assure proper facility design, and that equipment and utilities are performing as intended.

Process Performance Qualification (PPQ) Performance qualification protocol(s) Protocol considerations must go beyond just a particular number of batches made Criteria, including statistical criteria, that if met, leads to the conclusion that the process consistently produces quality product Can leverage previous experience and knowledge if the data is relevant to the commercial scale process Manufacturers must establish criteria and justify it. FDA cannot dictate as it is dependent on the type of product and process, its complexity and the risk associated with various steps or unit operations in the process. The main criteria isn’t how many commercial scale batches are made but rather what data are they collected and what are they doing with the data? How are they analysis the data? What should the result of this scientific, statistical where appropriate, analysis be? Generally a combination of in process measures and final product quality attributes.

Performance Qualification Typically will include: Commercial batches manufactured with the qualified utilities, facilities, production equipment, approved components, master production and control record, and trained production personnel in place. Usually run at target/nominal operating parameters within proven acceptable range or design space. Extensively tested, i.e., combination of samples analytically tested and increased process control monitoring beyond typical routine QC levels. These lots are generally intended for sale. Typically run at target conditions. Not in the early experimentation stage or first attempt at finding boundary conditions. Not to say they would make commercial lots prior with the purpose of doing just that but again each firm has to set up their design studies as appropriate.

Basis for Commercial Distribution “Success at this stage signals an important milestone in the product lifecycle. A manufacturer must successfully complete PPQ before commencing commercial distribution of the product. The decision to begin commercial distribution should be supported by data from commercial scale batches. Data from laboratory and pilot studies can provide additional assurance that the commercial manufacturing process performs as expected.” WHEN CAN YOU BEGIN COMMERCIAL DISTRIBUTION AND NOT BE VIOLATING CGMPS? Is the firm’s decision based on their assurance level. Decision must be deliberate, obvious and the firm takes responsibility for it.

Basis for Commercial Distribution “Each manufacturer should judge whether it has gained sufficient understanding to provide a high degree of assurance in its manufacturing process to justify release of the product.” WHEN CAN YOU BEGIN COMMERCIAL DISTRIBUTION AND NOT BE VIOLATING CGMPS? Is the firm’s decision based on their assurance level. Decision must be deliberate, obvious and the firm takes responsibility for it.

Stage 3: Continued Process Verification Process Validation during commercial manufacturing “An ongoing program must be established. The data collected should include relevant process trends and quality of incoming materials or components, in process material, and finished products. The data should be statistically trended and reviewed by trained personnel. The information collected should verify that the quality attributes are being appropriately controlled throughout the process.” Stage 3 of the lifecycle

Trend and Assess Data Evaluate periodically (at least annually1) to determine the need for changes in drug product specifications or manufacturing and control procedures Analyze data gathered from monitoring processes Incorporate statistical and/or quantitative measures where appropriate and feasible Study OOS and OOT (out of trend) data Assess impact of process and product changes over time Feedback into design stage for significant process shifts or changes 1 21 CFR 211.180(e) This is an important slide. Explain that stage three of PV really brings into play another GMP requirement: the periodic evaluation of data to determine the need for changes to specs, procedure or controls. As they gain manufacturing experience, they may learn more about the process and may need to change control strategy, revisit the design stage and do more experimentation or in some way adjust the process. 211.180(e) is related to monitoring and trending activities.

Stage 3: Continued Process Verification Pursue gaps in knowledge when discovered Follow up unexpected, unexplained results Revisit process design to improve current process robustness Conduct in depth root cause analyses when deviations occur. Further refine Control Strategy Be aggressive and timely in addressing issues

Changes to the Process Statutory and GMP References: FD&C Act Section 506A(b) “Manufacturing Changes” post marketing - requires validation of the effects of a change on the identity, strength, quality, purity and potency 21 CFR 211.100(a) “…written procedures, including any changes, shall be drafted, reviewed and approved…” 21 CFR 211. 180(e) – Annual review to determine whether changes in specifications or manufacturing or control procedures are needed” This slide points out where change is mentioned in the Statute and the regs. Changes need to be evaluated to assure they return the process to a state of control (if that’s why it’s being made) or don’t adversely impact the state of control.

Periodic Evaluation Re-validation – term is not used in the Process Validation Guidance Production phase (continual verification) monitoring will evaluate quality indicator data, changes and adverse trends and should be be used to periodically decide if new studies, e.g., conformances batches or other verification experiments, need to be done. The Guidance doesn’t address re-validation as Stage 3 essentially will cover that. If properly monitoring and proper change control, the depth and breadth of re-validation should be clear and it should be justified.

Continuous Processing Use of PAT systems to detect input and output variability of the material and react in real time to prevent sub-quality product Scale-up issues may not be as relevant as with batch manufacturing. Combination of real time detection of material attribute quality (particle size, uniformity) as well as process drift or change in performance (e.g. flow rates, power) Monitoring quality on a continuous basis. The principles expressed in the new draft Guidance would apply to continuous processing as well even though it and our regs are batch focused.

Summary The lifecycle approach to Process Validation links product/process development to the commercial manufacturing process, and maintains the process in a state-of-control during routine production. In conclusion, FDA is looking at PV as a product lifecycle activity. The new Guidance intends to link the process development, initial qualification and maintaining process control during routine production.

Do you have any questions? 谢谢 Xie xie. Thank you. 您是否有任何問題？ Nǐ yǒu shé me wèntí ma? Do you have any questions?