Critical Challenges to Implementing QbD: A Q&A with FDA

October 2, 2009
Angie Drakulich

Angie Drakulich was editorial director of Pharmaceutical Technology.

Pharmaceutical Technology, Pharmaceutical Technology-10-02-2009, Volume 33, Issue 10

Officials from the US Food and Drug Administration discuss best practices for applying quality-by-design concepts. This article contains bonus online-exclusive material.

The US Food and Drug Adminstration released its Pharmaceutical Quality for the 21st Century: A Risk-Based Approach report in 2004. Since that time, the concepts around quality by design (QbD) have been discussed in multiple meetings and regulatory guidances, including in three finalized guidelines by the International Conference on Harmonization (ICH) (1–3). Christine Moore, acting deputy director of the Office for New Drug Quality Assessment, and Joseph Famulare, deputy director of the Office of Compliance, both within FDA's Center for Drug Evaluation and Research (CDER), address some common industry questions about how to implement the manufacturing practice.

Core challenges

PharmTech» Controlling variability duringthe manufacturing process seems to be the biggest challenge facing industry with regard to QbD implementation because this approach involves different instrumentation and additional testing. And yet, process analytical technology (PAT) can offer industry great advantages. Why do you think there is so much pushback from industry with this issue, and what might it take for industry to accept the changed approach that FDA is promoting?

»Moore: I see a few major concerns of companies in shifting toward implementing PAT. First is the unavailability of the equipment in current pharmaceutical plants and the data-management systems needed for tracking, control, and knowledge management. Also, many manufacturers are unfamiliar with the equipment and its operation, which could lead to a steep initial learning curve and new needs for maintenance on the manufacturing floor. The second hurdle I perceive is cultural. The traditional mindset of the pharmaceutical industry often has been to put a validated process in place and not look at it again unless it breaks. PAT systems can allow for a continuous window on the operation which will give you much greater and richer information so that the manufacturer can more easily detect process drifts and correct them before they become problems. I think as more companies implement QbD and PAT approaches, the regulatory and business benefits will become more evident.

PharmTech» Pharmaceutical manufacturers also face QbD challenges with legacy products and scale-up. How does FDA intend to address these issues?

»Famulare: Firms have different motivations for applying new technologies to legacy products, including finding ways to gain better process control and efficiency. Commercial data, in conjunction with additional development work, can lead to anything from better process controls to seeing the need to update the application with new specifications for the product, and may or may not include the implementation of a design space. ICH Q10 Pharmaceutical Quality System (PQS) provides guidance in implementing these approaches during all phases of the product life cycle, including technology transfer and scale-up.

FDA continues to train compliance and investigative staff in advanced technologies, risk-management, and quality systems. CDER compliance technical experts lead and participate in preapproval and postapproval inspections on occasions where their expertise is needed and in the foreign arena, which has proven very helpful in looking at many of these opportunities. CDER's Office of Compliance (OC) has been instrumental in coordinating joint inspections with international colleagues [e.g., the European Medicines Agency] that have new technologies working and is open to preoperational visits where new technologies are put into place.

PharmTech» Some recent industry presentations have discussed the size of the design space and what's most effective. What's the regulatory perspective on the size of a design space and how close a company's process results come to the edges of the design space?

»Famulare: The value of design space is not its size—or where the process lies within the design space—but the process understanding it represents based upon good development and knowledge obtained during the commercial process over the product's life cycle when the design space is changed. This allow for a firm's self-management of change to foster innovation and continual improvement. In the past, many processes remained fixed based upon limited knowledge at the time of development and were limited to fixed processes prescribed in applications.

Design space will provide the opportunity to present a good understanding of the proposed process based on development and to innovate and improve the process. As described in ICH Q10, developing an effective control strategy based upon the most current product and process understanding should provide the opportunity for manufacturers to control variability and to produce a quality product minimizing deviations. In addition, FDA's draft process validation guidance addresses more clearly than ever before that validation is not a fixed activity but should be a life-cycle approach that can be adapted to work with or without design space and can be used with modern technologies that continually monitor or measure the process.

PharmTech» As you mentioned, ICH Q10 focuses on a life-cycle approach by adopting a quality system. What are some ways that industry can best apply this approach?

»Famulare: The life-cycle approach is at the heart of FDA's quality systems guidance published several years ago and has now been superseded internationally by ICH Q10 (see Figure 1). The challenge can be met by harnessing the data amassed in a commercial process to drive continual improvement.

Table I: Quality-by-design (QbD) programs in progress at the US Food and Drug Administration.

An important PQS element is the process-performance and product-quality monitoring system. That system will confirm whether predictions made in development have provided a good understanding of the process. QbD approaches to development will [also help a firm] to learn from the process in a knowledge management system (a PQS enabler).... The newer regulatory approaches derived from ICH Q8 Pharmaceutical Development, Q9 Quality Risk Management, and Q10 will allow companies to drive continual improvement. Processes will not be fixed by commitments made in applications that may not reflect an adequate understanding of the manufacturing process because of limited knowledge.

PharmTech» Why is it recommend that a company begin to think about QbD concepts at the end of Phase II? Is there ever an ending point for applying a quality-based approach (e.g., should it affect sales and marketing)?

»Famulare: Paraphrasing Qbd, it means a systematic approach to development, which is worthwhile at any phase of development. The intensity of certain development studies may increase at the end of Phase II (e.g., in deciding to use a design of experiment more extensively, whether or not to develop a design space for a particular unit operation, and whether certain technologies such as PAT are being considered for the commercial process). There will also be considerations as to what types of process controls will be used and consideration of real-time release testing (RTRT). ICH Q10 illustrates consideration for each phase of the product life cycle and the OC's Division of Manufacturing and Product Quality (DMPQ) has looked at proposals at various phases, especially when new facilities or manufacturing technologies are being considered. They can be contacted via subject matter experts listed on the website (www.fda.gov/Drugs/DevelopmentApprovalProcess/Manufacturing/default.html).

A PQS should extend throughout all parts of a firm's operations from preclinical, clinical, and marketing. This is a challenge CDER has seen firms meet by creating an opportunity to have lower deviations, fewer complaints, and to make better linkages from clinical to quality.

Lessons learned thus far

PharmTech» Now that the Chemistry, Manufacturing, and Controls (CMC) Pilot Program is winding down, will FDA be changing the way QbD-based applications for new drugs need to be submitted?

»Moore: Applications for QbD and traditional approaches are submitted in the same manner. From an administrative standpoint, it is helpful to us if the applicant mentions the QbD aspects or other unique features of the application in the cover letter.

PharmTech» What were some of the major lessons learned by FDA as a result of the CMC pilot?

»Moore: We learned how to make science- and risk-based regulatory decisions based on QbD data and proposed manufacturing approaches. Importantly, we gained experience dealing with different types of flexible regulatory approaches such as design spaces, in-process testing, and RTRT. We found that having a reviewer on Prior Approval Inspections was synergistic in that it benefitted both the review and inspection.

One of the major findings through the pilot was that quality risk-management and a pharmaceutical quality system are equally important to pharmaceutical development for the success of a QbD approach. As we get additional applications containing QbD approaches, we are continuing to refine our review practices, and are sorting out questions such as what detail of information should be in an application. We are further looking at how best to maintain our internal knowledge management and information-sharing across offices.

PharmTech» FDA has received more QbD-based submissions (12 new drug applications [NDAs], 18 investigational new drug applications [INDs], and 3 supplemental NDAs as of May 2009) than it did under the CMC Pilot Program. Does industry seem to have a hold on what FDA is looking for when it comes to QbD-based submissions?

»Moore: The early applications under the CMC pilot had most of their development complete by the time the pilot was announced. The applications were more like hybrids, containing both traditional and QbD approaches. Some of the newer applications that we are seeing were wholly developed using QbD approaches. Every application is different and we are still learning and coalescing our review approaches. For example, we don't have much experience in changes to approved design spaces.

Changing the pharma mindset

PharmTech» How long do you expect before QbD and its related tools and concepts become a natural practice in drug manufacturing?

»Moore: Many of the approaches of QbD (e.g., continual improvement) are a change to the traditional approach of putting a validated process in place and not touching it. I think that the pharmaceutical community has already made huge strides in advancing QbD concepts. Five years ago, practically no one had heard of design space, and formalized quality risk-management was rarely done. Today, we see a growing number of manufacturers incorporating these concepts in their day-to-day development practice and manufacturing operation.

PharmTech» One way to break down barriers to QbD implementation is to increase communication between industry and the regulatory authorities. Do you expect FDA to seek more interaction with firms during the review process or inspection process?

»Moore: We have already seen a greater level of interaction in many of the QbD applications, both within and outside of the pilot. As I often mention in my presentations, we encourage companies looking to implement QbD approaches to meet with us prior to their submission. Even though not all the details of the design space will be available yet, we find that an end-of-Phase-II meeting [as Mr. Famulare mentioned above] is a good time to start discussing the proposed approaches. Pre-NDA meetings are a good time to discuss more specifics of the content and format of a QbD application. For legacy products, a meeting can be requested to discuss proposed QbD approaches in a supplemental application.

PharmTech» Does FDA expect to see more communication and interaction between, for example, research and development (R&D) departments and quality-control departments?

»Famulare: Yes, these are really important links for successful implementation of a PQS approach within a company to relay not only information but knowledge from development to manufacturing with quality integrated throughout. This applies whether or not QbD approaches are used in product development and especially important to transmit commercial manufacturing information back to R&D. This will also provide for a more robust change-management system within a company to recognize when change can be made to foster continual improvement or in response to a corrective action/preventive action (CAPA). This is an important tenant of the ICH Q10 change-management system element.

It is also important for the regulator to have systems to transfer knowledge between review and inspection. The OC is developing a system of knowledge transfer to facilitate this communication.

PharmTech» There is an informal practice among industry to avoid providing FDA with too much detail out of fear that more information will lead to more questions, and perhaps more 483s. Can you address how the agency is trying to change this mindset?

»Moore: A discussion point in the FDA 2004 report on pharmaceutical quality was to focus on important, well-summarized information in the application, rather than voluminous raw data. I think quality risk-assessment is a good example of the additional important information in a QbD application. It is a change in mindset for companies to admit that they have risk. Of course, their processes have always had risk, and reviewers have always looked for the risks. Addressing these risks and their control in a formalized manner increases transparency and aids in understanding the firm's assurance of product quality.

PharmTech» Is the agency planning to increase its practical workshops or speaking engagements to help clarify the questions that still exist around QbD concepts and quality systems?

»Famulare: There are [still] questions about the regulatory implications [of ICH Q8, Q9, and Q10] and how to present some of these proposals to the agency and how investigators may look at these proposals. I believe that conferences, papers, and shared learning on good case studies will help answer some of these questions and relieve some of these concerns. Those regulators, including myself, and industry members involved in ICH quality initiatives realize that ICH Q8, Q9, and Q10 are forward-looking and that help is needed for implementation. Hence, the ICH Implementation Working Group (IWG), led by rappateur John Louis Robert, is looking to do that through papers and conferences beginning next year. Papers drafted will have input and guidance from I WG members. Any questions about manufacturing and various approaches to science- and risk-based approaches and CGMP interpretations can be directed to DMPQ. So far, we have looked at proposals for RTRT, PAT, and approaches to validation involving continuous manufacturing and verification.

References

1. ICH, Q8 (R2) Pharmaceutical Development (Geneva, Aug. 2009).

2. ICH, Q9 Quality Risk Management (Geneva, Nov. 2005).

3. ICH, Q10 Pharmaceutical Quality System (Portand, OR, June 2008).