To gain perspective on the implementation of quality by design (QbD) in sterile manufacturing, Pharmaceutical Technology spoke with key industry experts: Wolfgang Weikmann, vice-president quality assurance at Vetter Pharma-Fertigung GmbH & Co.KG; Hal Baseman COO and principal at ValSource LLC and chair elect of Parenteral Drug Association (PDA)Board of Directors and vice chair of PDA Science Advisory; and James E. Akers, president of Akers Kennedy and Associates.
Critical quality attributes
PharmTech: In implementing a QbD approach, what would you identify as the critical quality attributes (CQAs) in sterile manufacturing and aseptic processing?
Akers (Akers Kennedy and Associates): The critical quality attributes are related to the control and prevention of contamination, principally, microbiological, total particulate matter, and endotoxin although clearly the chemical purity and material uniformity issues that are a constant in pharmaceutical manufacturing apply to sterile and aseptic products as well.
Weikmann (Vetter): The potential critical quality attributes of a customer’s drug product in terms of aseptic processing are determined by the nature of the product and by the definition of its target product profile. Most common identified critical attributes refer to physical, chemical, and microbiological properties. We recognize sterility, bacterial endotoxin, and dose accuracy as three of the essential CQAs.
Baseman (ValSource): Identification and understanding of CQAs are essential to designing a process that establishes and maintains those CQAs. They can be broadly defined as purity, safety, strength, and identity of the product. Therefore, the CQAs associated with aseptic processing and those that need to be established and maintained by the process are:
- Purity: Those CQAs that assure that nothing is present in the product which should not be there (i.e., foreign material or cross contamination of product or chemical residuals).
- Safety: Those CQAs whose absence will likely cause harm to the patient, then they would include sterility, sterility assurance, lack of endoxin, lack of particulate contamination.
- Strength: Those CQAs that must be present to assure the effectiveness of the product, then they would include potency, active ingredient, and functionality
- Identity: Those CQAs that assure that the product is what it purports to be, then they would include correct and complete indication of product identity, lot number, content, and potency.
Critical process parameters
PharmTech: What measurement tools are typically used to measure critical process parameters and the resultant process inputs and outputs, including process analytical technology (PAT) tools, in sterile manufacturing?
Akers (Akers Kennedy and Associates): Many of the key parameters in aseptic and sterile manufacturing can be measured directly and in real time. Sterilization of components, depyrogenation of some materials, fill volumes, and in some cases, package sealing, can be measured in real time. Also, key environmental factors, such as total particulate air quality, pressure differentials, and air-handler operational characteristics can be measured in real time. In modern isolator/aseptic chamber aseptic manufacturing and in terminal sterilization, those parameters listed are the most critical.
Baseman (ValSource): Defining a quantifiable correlation between what can be measured and the intended outcome is particularly challenging for aseptic processing. Some tools and systems that can help better understand and design the process include:
- Air profile/smoke studies and material/personnel flow studies to establish first-air principles and their relationship to potentially microbial contamination. These tools are essential for the design and assurance of the process control strategies.
- Environmental monitoring, including continuous nonviable and viable monitoring, to monitor performance of the process.
- Risk assessments to better understand the relationship and impact of process control strategies on CQAs and critical process parameters.
- Human factors and ergonomics to help create processes that minimize the effect of human performance and fatigue.
- Differential pressure and airflow and velocity to maintain cleanliness of the cleanroom environment.
- Temperature and humidity monitoring to reduce the potential for microbiological growth in the cleanroom.
- Gown and personnel monitoring to assure control of environmental conditions.
There is, however, a direct and measurable correlation between sterilization parameters, including:
- Time, temperature, and pressure monitoring of steam-sterilization cycles to assure control of microbiological contamination of product contact parts and surfaces.
- Some of the aseptic processes, such as lyophilization, may use PAT in their recipe management for process consistency and control outcome CQAs.
- Filtration integrity testing to assure integrity and effectiveness of product sterilization methods.
PharmTech: How is equipment design and operation modified when implementing QbD in sterile and aseptic manufacturing?
Baseman (ValSource): Air profile/smoke studies can be used to determine the most optimal shape and configuration for aseptic processing equipment. Human factors can be used to determine and improve interventions. Risk-based evaluation of interventions and use of automation can be used to reduce the impact and risk of interventions on process variation and product quality.
Weikmann (Vetter): Looking first at equipment design, we feel strongly that the design review process has to be mentioned as an essential tool. Design review is a decisive milestone within the development process for a single piece of equipment. It is performed through the adjustment of an initially created design with its particular requirements. This is important to control the results of the former activities, as well as to identify and solve any existing problems prior to the start of the drug-manufacturing process. We primarily execute a design review within the construction phase for each cGMP relevant piece of major process equipment. An example element of this design review is conducting mock-up studies.
Looking at the operations side, serving multiple customers with different product and process specific requirements, our equipment has to be set-up highly flexible. Such an arrangement offers the benefit that, based on the product specific requirements as well as based on the performed pump studies within the QbD approach, the specific equipment components, such as the most adequate product and process-fitting filling pumps, can be used right away for subsequent production.
Akers (Akers Kennedy and Associates): In the projects I’ve worked on over the past 20 years, I don’t think current emphasis on QbD has altered what we do because we’ve always been focused on arriving at the most effective, efficient, and safe design for a process. I think because of the current regulatory emphasis, we use the description QbD or as we say sterility by design more than in the past. So, there is some additional focus in thinking and speaking in QbD terms, particularly among some clients.
PharmTech: What are the challenges in applying QbD in sterile manufacturing?
Baseman (ValSource): [The challenges are] the lack of quantifiable correlation between what we can measure and monitor, and the desired outcome or the undesired outcome. Also, one must be aware of residual risk and unintended consequences as a result of changes made to address risk and improve process. Each action taken to resolve or address a process risk represents a change that may have the potential to add a new risk or unintended consequence. These new risks must be evaluated and addressed where possible.
Akers (Akers Kennedy and Associates): QbD is not really a new requirement in my experience. I believe in sterile manufacturing; we’ve done QbD before it was formally termed QbD. I believe much of what we now do as QbD was once part of our design qualification activities. The design parameters for sterile production facilities in terms of cleanrooms and sanitary designs have been well defined for years.
Weikmann (Vetter): In our experience, the main challenges in applying a QbD approach are due to the intensity of time, cost, and resources in the early-development stage. Another significant challenge is the high degree of expertise that must be in place to run a proper QbD approach.
Regarding timing, a much longer timetable has to be included. As for costs, the higher costs are due to the variety of additionally performed studies. And higher resource management is due to equipment, different materials as well as staff.
In summation, it is always a product-dependent balancing act between the risks and consequences on one hand, and the cost during the development stage on the other. Even if challenges as previously described appear in the early stages, the gained knowledge will likely result in an economic benefit in the long-term commercial lifecycle.
Benefits of QbD
PharmTech: What additional understanding is gained through a QbD approach compared to a traditional approach?
Weikmann (Vetter): The most important understanding gained by running an operation through a QbD approach versus the more traditional approach is to realize the necessary requirements of a customer´s product already in an early-development stage to accommodate the product target profile. These requirements are not limited to product quality only. They include the process performance as well as the applied systems and the environment, for example, optimized and reduced lyophilization process cycles.
Based on various studies, differing possible cases are evaluated in detail. This examination ultimately leads to a better understanding of the process variations, thus resulting in an acceptable design space. Consequently, using a QbD approach offers the possibility of gaining better process knowledge, resulting in an improved robustness of the entire process.
Baseman (ValSource): A better understanding of process variables and impact on risk to product quality [is gained], which will aid to design of the process and process control strategies.
Akers (Akers Kennedy and Associates): Actually, Joseph M. Juran coined the phrase ‘quality by design’ in the early 1990s, and these ideas were widely adopted in aseptic processing before QbD re-entered the industry lexicon a few years ago. I would estimate that 95% of aseptic product safety in 2013 emanates directly from the proper design of the equipment and the environment.
PharmTech: How does QbD improve quality assurance and how can it mitigate problems in manufacturing?
Weikmann (Vetter): Through a greater importance given to process development activities and the resulting large number of development studies performed within a QbD approach, a better process understanding is gained. Within the risk-management approach, risks are already identified in an early stage. Consequently, risk mitigation activities can be developed, and a risk-based control strategy implemented from the very beginning of the commercial stage, forming the basis for improvement of quality assurance and quality oversight.
Akers (Akers Kennedy and Associates): I can’t point to a single specific area where I believe QbD has had an impact on sterile product quality assurance. Sterile product manufacturing has always been and will continue to be a detail-oriented endeavor. QbD may have increased awareness of the criticality of getting the facility and process designs right. Maybe we have a better appreciation for the primacy of good engineering and scientific process development in sterile products manufacturing operations.
Baseman (ValSource): It can improve process understanding, including the identification and control of process variables. This can lead to improved process and less risk.
PharmTech: How do QbD practices build on or complement approaches taken with respect to parametric release and related terminal sterilization parameters needed for release?
Akers (Akers Kennedy and Associates): Perhaps QbD, as previously noted, will strengthen the focus on process design and control. However, arriving at a process suitable for consideration for parametric release has always required very thorough science and engineering. I personally believe that terminally sterilized products, given their robustness and reliability, should be parametrically released more often than not. The alternative to parametric release is the sterility test, which has limitations in terms of microbiological detection and sensitivity as well as in the area of sampling statistics.
Baseman (ValSource): Understanding the relationship between those parameters we can control, monitor, and measure, and the desired outcome or CQA is essential to designing effective processes. It may not be enough to know which parameters effect product quality and process performance; we must also know to what extent they do this and how well they are controlled.
PharmTech: How is technology transfer improved under a QbD approach?
Weikmann (Vetter): Technology transfer indeed is one area that is especially improved through a QbD approach. It leads to improved possibilities to monitor and evaluate the process. Becoming familiar with a drug product´s frame and its limits through conducting numerous tech runs of varying versions allows one to define the product´s specification in which the process is running robustly.
A day-to-day business example to showcase the improvement of tech transfer through QbD for customers is by demonstrating the lyophilization cycle development of a customer’s product. For example, by using a QbD approach with gaining the corresponding data based on numerous performed lyophilization tests, we realized a significant reduction of the lyophilization cycle for a particular customer product. This reduction resulted in saving the customer a high amount of capital, primarily within the large-scale commercial production.
Akers (Akers Kennedy and Associates): I honestly don’t see much change. I’ve been fortunate enough to work with firms who always took the development of user-requirement specifications through detailed proof of principal testing very seriously.
Baseman (ValSource): It can result in a better understanding of the process and process variables, including which process steps require control--control measures that need to be included in tech transfer.