Paradigm Shifts in Pharmaceutical Quality Systems
A Historical Review and Future Perspectives [Part 2]
Implementation of Process Validation
The concept of validation has its roots in the space technology developed during the U.S. Apollo program (1960s). Launch rockets are single-use. The concept of sampling inspection—where one might say, “Let’s test-launch it; since it worked, we’ll launch it again”—simply does not apply. Therefore, NASA rigorously applied the principles of reliability engineering—the idea that “if the design and manufacturing processes are flawless, the final product must also be flawless”—and established the prototype for the concept of validation.
FDA officials wondered, “Why isn’t the same level of precision in quality assurance found in the rocket industry being applied to the pharmaceutical industry, where human lives are at stake?” and introduced the NASA-derived approach of “building quality into the process” into the world of GMP. It can be said that validation is a technology “adapted from the pharmaceutical world to the quality assurance techniques cultivated in NASA’s Apollo program—an environment where failure was not an option.” Understanding this background should help explain why validation places such a strong emphasis on “documentation” and “advance planning.”
The definition of process validation issued in 1987 was “the establishment of evidence that provides a high degree of assurance that a process can consistently produce products that meet specified standards and quality characteristics.” But what does “consistently produce” actually mean? Since the definition provided only a conceptual framework without specifying concrete methods for validating pharmaceutical manufacturing processes, it caused considerable confusion among regulatory authorities and pharmaceutical companies worldwide.
The FDA must have been quite troubled by this as well; in 1993, the FDA specifically presented “the successful production of three consecutive batches at actual production scale” as a requirement for completing validation. This became widely recognized within the industry, and the “three-batch rule” took root worldwide as the “absolute standard for passing FDA inspections.” Lacking scientific basis, the misconception that “three successful runs guarantee lifelong security” spread. As a result, companies were unable to respond to fluctuations occurring after market release and were plagued by situations where “defective products emerged despite validation.” Furthermore, once a process was validated, even the slightest change in procedure required “re-validation,” leading to a perverse situation where production sites hesitated to improve manufacturing methods.
FDA revised its guidelines on process validation to address this situation as an activity involving the collection and evaluation of necessary data throughout the product lifecycle in 2011. In other words, FDA clarified that process validation is not a one-time activity, such as the successful completion of three consecutive batches, but rather an ongoing process of continuous improvement throughout the lifecycle of a pharmaceutical product.
That is, process validation is a series of activities that begins with Stage 1: Process Design, proceeds through Stage 2: Process Performance Qualification (PPQ) at actual production scale, and culminates in Stage 3: Ongoing Verification during commercial manufacturing; this means that process validation activities never end as long as the drug remains on the market.
Process validation now consists largely of qualification (Stages 1 and 2); once qualification is complete, the process involves continuing verification while collecting data to improve the process, and is no longer a one-time activity like the previous “three-batch rule.” This was indeed a major change that threw the industry into disarray.
In other words, PPQ has shifted from the unscientific three-batch-rule to a requirement to demonstrate the necessary number of batches through scientific justification. The primary basis for determining the number of batches is the qualification data accumulated during the Stage 1 development phase and the scale-up phase. Based on this data, we predict “variability” in commercial production and determine the required number of batches through risk assessment and statistical analysis. There is a stark difference in the level of quality assurance and the persuasiveness during inspections between the old attitude of “We made three batches because that’s the rule” and the new approach of “After statistically calculating variability and evaluating risk, we concluded that three batches are sufficient to demonstrate the reliability of our product.” It can be said that this reflects how deeply the company understands its own processes.
If process qualification is thorough, a single PPQ may suffice. Ultimately, the essence of process validation lies in process qualification, while PPQ and ongoing verification serve merely to confirm it.
At the time, I was conducting scale-up studies for a synthetic API. Since I firmly believed that as long as we could determine the scale-up factors for the synthesis process, we would undoubtedly be able to produce the product, I thought the “three-batch rule” for process validation was unreasonable. At least, for synthetic API processes, I believed that if there was a scientific basis for the process control parameters, verification of a single batch would be sufficient.
In particular, when the final process step involves purification by recrystallization, I believe that excessive validation is unnecessary because a 100% inspection can be performed to ensure that the crude crystals are completely dissolved and the mixture is homogeneous. Of course, documentation is required by regulations, but as long as there is scientific justification, the PPQ should be straightforward.
However, the situation is different in the world of pharmaceutical formulation. For example, there is no scientifically validated method for uniformly mixing two powders with different properties. Segregation may occur during mixing, and extending the mixing time does not necessarily result in greater uniformity. Furthermore, there is no scientifically validated sampling method to confirm that the mixture has become uniform. Thus, when a process cannot be fully qualified, simply achieving three consecutive successful batches is insufficient to complete PPQ.
The number of batches required for PPQ must be determined based on the level of qualification of that process.
Process validation, which originated in rocket science, caused a sensation in the pharmaceutical industry. However, it ultimately settled into a very ordinary manufacturing activity: qualifying processes, operating them at commercial scale to verify performance, managing them on a daily basis, and making improvements when problems arise.
While questions regarding analytical method validation or cleaning procedure validation can be answered with a simple “Yes” or “No,” process validation has become a unique case in that the only possible answer is that it is always in progress. For the sake of convenience, I believe most people interpret PPQ as process validation. In any case, it cannot be denied that confusion arose because the FDA changed its definition of process validation from a one-time activity to a continuous activity spanning the entire lifecycle of a drug.
Author
Nakao started his career in 1976 at Tanabe Seiyaku Co., Ltd. After carrying out research in synthetic organic chemistry and process chemistry (industrialization) at the company, he successively took the positions of quality assurance manager, director and general manager of production, and managing director in charge of management planning. After Tanabe merged with Mitsubishi to become Mitsubishi Tanabe Pharma Corporation, Nakao became Managing Executive Officer and Chief Director of Medicine Manufacturing. Through his experience in dealing with FDA inspections and activities at the Parenteral Drug Association (PDA), Nakao became convinced that GMP is an “exact science.”
Finding that his beliefs were reflected in CM Plus Corporation’s enthusiasm for improving GMP education, Nakao joined CM Plus in 2011 as a Director and be promoted as executive vice-president in 2012. Fellow since July 2017.
About GMP Platform (https://www.gmp-platform.com/)
*If you are using Google Chrome, right-click and select “Google Translate” on the article.