Improving Prospective Product Development Methods Derived from Management-Based Regulatory Guidelines

Peer-Reviewed: 
Submitted: Jan. 3, 2019. Accepted: Jan. 28, 2019

To rapidly achieve high-quality pharmaceutical manufacturing processes, the industry must focus on developing prospective, management-based approaches instead of retrospective performance-based measures that are too slow to drive the necessary innovation and continuous improvement. FDA’s voluntary consensus standards initiative may provide the opportunity for establishing and evolving important prospective product development methods.

Regulatory agencies, including FDA, have switched from a management-based to a performance-based regulatory strategy. The stated reason for the change is “to give the industry enough flexibility to manage and improve quality on its own” in hopes of a several order-of-magnitude advance in performance using continuous improvement (1). The need for improving performance is caused by the industry’s ongoing failure to reliably supply patients with high-quality products resulting in many shortages and recalls due to industry-wide manufacturing quality levels around two to three-sigma (two-sigma is roughly a 30% failure rate) (1,2). In simple terms, a poorly performing industry is being told by regulatory agencies to “fix thyself.”

A case can be made that the shift has left the industry adrift in the rough seas of rapidly advancing medical technology without the necessary methods to keep up. The shift in regulatory strategy places a huge, but necessary burden on the industry to develop their own sophisticated, prospective, management-based methods required to achieve Six Sigma quality levels (roughly three failures/million attempts) typically found in many other industries while developing increasingly complex products and therapies. 

Management-based approaches must be prospective if they are to be effective for developing and manufacturing future high-quality products. Performance-based approaches, particularly regulations, are retrospective measures of quality. Unless the management-based methods are effective, feedback from performance-based methods that depend on initial or past failures to improve performance will be slow to improve the industry’s performance. For pharmaceuticals, high performance must be achieved from the very beginning using prospectively developed control strategies. Further, these better methods must be developed and improved in collaboration with regulatory agencies to ensure the methods will have a perceived high likelihood of actually producing the initial results necessary to support regulatory actions for licensure and commercial manufacturing. The collaboration must also ensure that the methods’ results are consistent with existing regulations. A possible method for the industry collaborating with regulatory agencies is being developed around the voluntary consensus standards (VCS) draft guidance (3). 

The best path for developing better industrial methods is likely to expand and supplement the existing management-based regulations developed before the shift in regulatory strategy. Old regulatory guidelines have been carefully written to avoid prescribing how approaches should be executed from a fear of establishing standardized methods that might prevent better approaches from being developed through continuous improvement. Only the industry, with the assistance of regulatory agencies, can prospectively develop and, most importantly, continue to evolve new and better methods and approaches to achieve excellence for ever-increasing product and manufacturing challenges.

This article reviews and suggests improvements to a few of the important management-based regulatory guidelines that currently drive pharmaceutical development and manufacturing. 

ICH Q8 (R2)-Pharmaceutical Development

The ICH Q8 guidance contains important foundational concepts for prospectively managing the development of high-quality processes necessary for making high-quality products. However, these concepts are defined vaguely as “what” and “why” and must be expanded or industrialized, perhaps using the VCS route, to achieve their intended goal of guiding the industry. For example, quality-by-design (QbD) is defined in ICH Q8 as:

A systematic approach to development that begins with predefined objectives and emphasizes product and process understanding and process control, based on sound science and quality risk management (4).

This mechanistic-free definition has led to widely varying interpretations and implementations of QbD. A study commissioned by FDA concluded that the vague definition has led to many problems within the industry and regulatory agencies on how QbD can and should be used (5). Insights into the regulatory goals of QbD have been described, but again these goals focus mostly on a regulatory perspective of past performance (setting specifications, process capability, and change control) rather than methods for achieving excellence from the very beginning (6).

A more mechanistic description has been proposed as working-quality by design (wQbD) by the following definition:  

To achieve a well-defined goal during the design stage, iteratively apply science and engineering methods to anticipate, identify, understand, and resolve problems that will be encountered during testing, operating, and verifying the goal over its entire lifecycle (7).

This definition describes the use of thought experiments combined with testing mathematical and experimental models within a larger design algorithm, such as the design stage of the lifecycle process development and validation paradigm described in FDA’s 2011 process validation guidance (8). As will be discussed, wQbD drives the systematic use of prospective quality risk management (QRM), operations research (OR), and design of experiments (DoE) methods to build effective control strategies for controlling risks and improving process performance. 

The results of the wQbD methods must be assembled and stored in a comprehensive process model for internal and regulatory communications. ICH Q8 defines the all-important process model as the design space described by the definition:

Design space. The multidimensional combination of interactions of input variable (e.g., material attributes) and process parameters that have been demonstrated to provide assurance of quality. 

Working within the design space is not considered as a change. Movement out of the design space is considered to be a change and would normally initiate a regulatory post approval change process. Design space is proposed by the applicant and is subject to regulatory assessment and approval (4).

The first sentence is the functional definition while the remainder are three increasingly significant policy statements intended to provide emphasis and guidance. Again, no mechanism is provided. The three policy statements provide a tantalizing look into the possibilities of sophisticated management-based regulations. Implementation of these policies, however, remains unlikely given the regulatory shift and the recent draft of ICH Q12 that describes comprehensive change control (9).

The regulatory design space is defined primarily in terms of the following two ICH Q8 definitions:  

Critical quality attribute (CQA). A physical, chemical, biological, or microbiological property or characteristic that should be within an appropriate limit, range, or distribution to ensure the desired product quality. 

Critical process parameter (CPP). A process parameter whose variability has an impact on a CQA and therefore should be monitored and controlled to ensure the process produces the desired quality (4).

Defining the design space in terms of the CQAs and CPPs is sufficient for regulating the performance of a process, but when used to design and clearly define manufacturing processes, the limited definitions have led to a great deal of confusion and uncertainty. Much clearer definitions of process inputs and outputs are required to build control strategies (10). 

To more completely describe the performance of a manufacturing process, both the product and the process’s behavior during product manufacturing are required. An alternative method of describing process outputs, previously proposed by the author (10,11), are:

CQA. Use the ICH Q8 definition, except it refers to the product material produced by each unit operation in the manufacturing process. In the case of the final process, the output CQAs would describe the final product. 

Critical process response (CPR). A process output that measures or indicates the performance of the process (e.g., temperature, pressure, viability, yield, etc.) that results in all the product’s CQAs. For complex biopharmaceutical products, measuring and controlling CPRs is required to control both unknown or unmeasurable CQAs as well as the known and measurable CQAs. If appropriate, a measurable CQA can be used as a CPR (10). A CPR is an output parameter using the above CPP regulatory definition. Clearly describing the performance of a process as a CPR focuses the development activities on the important aspects of controlling the process’s behavior to achieve maximum control over product quality. 

With respect to defining input parameters, the regulatory definition of the various input CCPs should be divided into the proposed following three categories (10,11):

Critical operating parameters (COPs). Input parameters used to define and manipulate the process during development and operation. Examples of COPs include media composition, buffer concentrations, mixing rates, gas addition rates, etc. COPs used to control the performance of the process in real-time based on measuring CPRs in a feedback control loop are called critical control parameters (CCPs). 

Critical material parameters (CMPs). Material attributes of the raw material input used by the process to make a product. CMPs are the same as a critical material attributes (CMAs). Thus, a series of processes are connected by the transfer of output material described by CQAs from one process that become the input material described by CMP/CMAs to the next process (6,11). 

Critical equipment parameters (CEPs). These parameters (sometimes called critical design parameters) define the equipment. Examples of CEPs include volumes, materials of construction, agitator type, heat transfer area, and so on. These parameters are part of the equipment’s user requirements specifications (URS) when the equipment is designed, selected, and acquired during engineering. 

When designing process systems, separating the three types of input parameters is important because they play different roles in determining the performance of the process. These parameters also contribute different risk input threats to the performance of the process and are controlled using different control strategies (10). Using the expanded definitions results in a well-structured design space (11). 

ICH Q9-Quality Risk Management (QRM)

ICH Q9-QRM is the regulatory guideline most urgently in need of rapid improvement and ongoing evolution (12). Too many times, regulatory guidance defaults to “do a risk analysis” to solve a problem or define a control strategy. Although ICH Q9 clearly states the “what and why,” the methods included in the document do not provide a viable foundation for the “how.” The methods described in ICH Q9 focus on process analysis (e.g., wishbone diagrams, failure mode and effect analysis, hazard analysis critical control point). Although an execution risk’s severity from a process can be described easily in a variety of ways (e.g., cost, supply impact), the highly subjective element of uncertainty, or likelihood of the risk’s realization, is not properly addressed by these methods. Improper handling of uncertainty frequently leads to confusion and misleading results (13,14). 

The failure to deal quantitatively with uncertainty often leads to the risk analysis falling prey to the precautionary principle, which can be defined here as an overemphasis on the impact of the risk’s severity without properly considering the risk’s probability as a result of inadequate information or poor analysis methods of the likelihood of the risk’s realization (15). The industry desperately needs effective quality risk management methods that are prospective, theoretically consistent with using probabilities to describe uncertainty, mathematically sound, easy to use by all practitioners, and expandable to an appropriate level of detail for analysis and control based on the risk’s severity (16–18).

FDA’s 2011 process validation guidelines

Arguably the most important guidance for developing control systems using QRM is FDA’s 2011 process validation guidelines (8). The document provides an important paradigm for developing and manufacturing products. The paradigm can be used to develop virtually any process from procedures or facility designs to unit operations (10,16,18). The guidance is unusual because it is very close to a prescriptive “how” paradigm. The approach is aligned with the four basic questions for any undertaking: “Do what?”, “How to do it?”, “Will it work?”, and “Did it work?” (10). Any analysis method that does not address all four questions is incomplete. If the four questions are correctly asked and answered, particularly in a systematic, documented fashion using wQbD to answer the second question, success is highly likely. 

FDA’s 2011 process validation guidelines describe three stages (8):

Stage 1: Process design. The commercial manufacturing process is defined during this stage based on knowledge gained through development and scale-up activities. 

Stage 2: Process qualification. During this stage, the process design is evaluated to determine if the process is capable of reproducible commercial manufacturing. 

Stage 3: Continued process verification. Ongoing assurance is gained during routine production that the process remains in a state of control. 

Although the above stages provide a good foundation, they do not adequately cover the four basic lifecycle questions necessary to provide a complete development paradigm for building excellence. They also lack mechanisms for their execution. A more complete lifecycle can be divided into planning stages for prospectively building the process along with its control strategies, followed by execution stages for retrospectively assuring initial and ongoing performance based on data. The execution stages provide the foundation for the performance-based regulations and are most effective when the process’s established conditions and quality metrics are intrinsically defined and incorporated during the planning stages. 

Briefly, the expanded process validation lifecycle paradigm can be described as planning stages (Stage 0 and 1) for prospectively building and documenting the foundation for manufacturing and execution stages (Stage 2 and 3) for retrospectively proving and documenting initial and ongoing success based on predefined comprehensive performance criteria described in the well-structured design space. The stages are described as follows:

Stage 0: Define. Establish clear product and process requirements, goals, and objectives to define success during the execution stages.

Stage 1: Design. Using wQbD, DoE, and prospective QRM to systematically develop control strategies and performance criteria within a well-structured design space to test, qualify, operate, control, and verify all process inputs and outputs required to manufacturing high quality products. 

Stage 2: Test and qualify. Execute a documented prospective testing program that ensures the process will work during Stage 3.

Stage 3: Operate, control, and verify. Operate and control the process using the control strategies defined and designed in the planning stages while assuring and documenting successful performance using prospectively defined performance criteria (10). 

The key to achieving very high product quality is to prospectively build a foundation during the planning stages using wQbD and QRM for the retrospective assurance of performance during execution by building highly effective control strategies, including integral measures of performance.

Summary

If the industry and regulatory agencies are to achieve FDA’s 21st century goals, the focus must be on developing and continuously improving prospective methods for developing new products and their manufacturing control strategies (19). While the existing management-based regulatory guidelines (ICH Q8 and Q9 and FDA 2011 process validation, etc.) are a good start, they must be improved significantly into straightforward methods necessary to achieve Six Sigma levels of performance to reliably supply all patients with high-quality products. Building and working with such methods will require a great deal of courage on the part of the pharmaceutical industry. Regulatory agencies and the pharmaceutical industry will have to work together, possibly using voluntary consensus standards, to build the methods along with the collective trust to efficiently develop, launch, and manufacture the many challenging products of the future. 

References

1.    L. X. Yu and M. Kopcha, Int. J. Pharm. 528 (1–2) 354–9 (2017).
2.    J. Wechsler, Pharm. Tech. 42 (12) 14-15 (2018).
3.    FDA, Draft Guidance for Industry, CDER’s Program for the Recognition of Voluntary Consensus Standards Related to Pharmaceutical Quality (February 2019).
4.    FDA, Guidance for Industry: Q8 (R2) Pharmaceutical Development, Rev. 2 (November 2009).
5.    FDA, Final Deliverable for FDA Understanding Challenges to QbD Project, Understanding Challenges to Quality by Design (Dec. 18, 2009).
6.    L. X. Yu, et al. AAPS J. 16 (4) 771–783 (2014).
7.    M. F. Witcher, BioProcess J. 13 (1) 6–11 (2014).
8.    FDA, Guidance for industry: Process Validation: General Principles and Practices, Rev. 1, (January 2011).
9.    ICH, Q12 Technical and Regulatory Considerations for Pharmaceutical Product Lifecycle Management Core Guideline, Draft version (endorsed on Nov. 16, 2017).
10.  M. F. Witcher, BioProcess J. online, DOI:10.12665/J17OA.Witcher.0416 (Sept. 5, 2018).
11.  M. F. Witcher, BioProcess J. online, DOI:10.12665/J132.Witcher (Summer 2014).
12.  FDA, Guidance for Industry: Q9 Quality Risk Management (June 2006).
13.  D. W. Hubbard, The Failure of Risk Management: Why It’s Broken How to Fix It (Wiley & Sons, 2009).
14.  L.A. Cox, D. Babayev, and W. Huber; Risk Analysis 25 (3)  651-662 (2005).
15.  C. R. Sunstein, Laws of Fear–Beyond the Precautionary Principle (Cambridge University Press, 2005).
16.  M. F. Witcher, BioProcess J. online DOI/10.12665/J16OA.Witcher (Sept. 25, 2017).
17.  M. F. Witcher, “Understanding and Analyzing the Uncertainty of Pharmaceutical Development and Manufacturing Execution Risks using a Prospective Causal Risk Model” (submitted to BioProcess J. 2019).
18.  M. F. Witcher, “Building Control Strategies to manage Pharmaceutical Development and Manufacturing Execution Risks using an Expanded Lifecycle Process Development and Validation Paradigm” (submitted to BioProcess J. 2019).
19. FDA, Pharmaceutical CGMPs for the 21st Century-a Risk Based Approach (September, 2004). 

About the author

Mark F. Witcher, PhD is senior consultant at Brevitas Consulting, witchermf@aol.com.

Article Details

Pharmaceutical Technology
Vol. 43, No. 7
July 2019
Pages: 34–37

Citation 

When referring to this article, please cite it as M. Witcher, “Improving Prospective Product Development Methods Derived from Management-Based Regulatory Guidelines," Pharmaceutical Technology 43 (7) 2019.