NIIMBL-Facilitated Active Listening Meeting between Industry and FDA Identifies Common Challenges for Adoption of New Biopharmaceutical Manufacturing Technologies

3.2.1. Interaction between Industry and the FDA (11).

Individuals at the FDA often serve many different roles, which can include that of a product reviewer and that of a regulatory scientist. Survey participants felt that it would be helpful to have more engagement with the FDA in their role of regulatory scientists, using mechanisms that do not compromise their role as regulators.

Transparency (10). It was reflected that a better understanding of the Agency’s philosophy, context for inquiries, and opportunities for earlier and more informal engagement could improve the overall quality and speed of submissions.

1. Shared learnings for implementing new technologies—A mechanism by which the FDA could share common gaps in submissions and contribute to discussions regarding advanced technologies or strategies that have been successful would reduce industry-perceived risks. These could take place in the context of public workshops or in a blinded fashion through a consortium or other mechanism.

2. Context for reviewer questions—There is a desire to better understand the context in which a particular review question is generated. Precompetitive sharing in an open forum that allows discussion to build a shared understanding of the context for reviewers’ questions would be very helpful for both understanding and adequately addressing the question.

3. Benchmarking and data sharing—FDA staff have access to an abundance of data and it was understood that there are sensitivities in sharing data, even if it is anonymized appropriately. It was reflected that talks given by FDA staff at conferences, particularly those that use case studies, are extremely helpful. The most helpful learnings are gleaned from a statement of the problem, how the FDA viewed the problem, and how they approached solving the problem. However, the potential for public scrutiny of FDA presentations and the disconnect that occurs after the meeting are limitations to collaboration.

4. Volume of data shared in regulatory filings—There were questions around the right amount of information to share with regulatory authorities and a sense that there is a trend toward an ever-increasing volume of data, not all of which may be useful or necessary even from the sponsor perspective. Further clarity around what information is needed in a dossier versus what can be managed by internal quality management systems would be helpful.

Collaboration (10). There was a sentiment that both industry and the Agency would benefit from increased collaboration to facilitate industry taking more business risk in implementing new technologies. Public–private partnerships are a potential mechanism for such collaboration. One idea is for industry and the Agency to use a consortium such as NIIMBL to engage with scientists broadly in forums that allow the participation by the Agency as regulatory scientists (as opposed to regulators), on a NIIMBL platform process.

1. Platform processes—Platform processes run by consortia such as NIIMBL could also be used to generate comparisons between emerging technologies and traditional technologies. It could be a forum for technology-oriented and non-product-specific discussions about what data might be appropriate to include in a regulatory filing. This information could be made available in the form of peer-reviewed publications.

2. Public workshops—Another avenue to increase collaboration could be more FDA-driven public workshops on new technologies. These could provide an opportunity for early dialog between regulators and industry members, ideally with the shared goal of expediting emerging technology implementation to improve the manufacturing of biopharmaceuticals. Some example topics raised included: multi-attribute methods, real-time release, traditional stability studies (particularly use of accelerated degradation for prediction and modeling), continuous manufacturing, and rapid testing (microbial, sterility, mycoplasma, etc.).

Early interaction and the Emerging Technologies Team (ETT) (9). Feedback about the CDER Emerging Technologies Team (ETT) was positive; however, nearly all the experience reflected was with small molecules rather than biotechnology products. Positive aspects of the ETT include the willingness to openly engage in a dialog at the conceptual level and having a point of scientific engagement with the Agency outside of the context of the review of a particular submission. It was reflected that it is helpful to have an internal expert on a technology that can be the bridge to the individual reviewer or inspector, which helps build experience and improve comfort levels. Engagement with the ETT was often viewed as a formal FDA interaction and the amount of data required for the interaction perceived as prohibitive.

1. Early engagement—Although the ETT is a valuable and perhaps underutilized mechanism, there is industry desire for more early engagement outside of the context of a particular filing. These early interactions could take the form of a forum to ask questions, perform benchmarking, or to find out what general approaches have been successful in the past.

2. Informal interactions—It was stated that there was a general understanding that these informal interactions would complement, but would not replace the need for more formal approaches (e.g., to expedite regulatory pathways for post-approval changes). There is a desire for more options for how to provide information and data to the Agency for post-approval changes, driven by a desire to improve at the formal submission stage.

Case studies (9). Case studies, such as the A-Mab case study, provided a safe space for individuals to share their thought processes and to learn how others in the industry and at the Agency thought about different issues. Many viewed the outcomes of the A-Mab case study as valuable guiding principles for quality by design (QbD) and risk assessment and as important for establishing a common language and understanding.

However, for the A-Mab case study, there was a sense of a disconnect between the outcomes published and what happened in practice with respect to QbD. Future endeavors would benefit from a mindful approach to minimizing the disparity between expectations and reality and tempering expectations.

Future case studies could generate non-product-specific dialog about new technologies between industry and the Agency. Pitfalls could be avoided with clear communication throughout the exercise, such as by choosing a topic that addresses a relative and immediate need, and by outlining clear goals and objectives of the exercise. An example topic includes a mock early-stage filing that uses new technologies for a hypothetical molecule that implements ICH Q12 (1) concepts in a filing, specifically analytical comparability using a new technology. Another example would be “A-Mab 2.0”, which would use the same hypothetical molecule but focus on a process based entirely on new manufacturing technologies.

3.2.2. Changes to Approved Manufacturing Processes (11).

Approaches to comparability (10). Demonstrating comparability often requires clinical data, and collecting this additional data can be expensive and time consuming. In contrast, it was also reflected that the expectations for analytical vs. in vivo vs. clinical data were very clear and these decisions were managed internally to be consistent with Agency expectations. Implementing new technology and deciding the best path to establishing comparability often includes the business risk of potentially needing to collect more clinical data.

Implementing a new technology might improve product quality in known ways, but a change to product quality is still a change to the product and would require clinical data.

A question was raised: Is comparability an adequate tool to bridge an existing technology with a new one? It may not be possible to truly compare Old Technology A to New Technology B, but rather a new assessment of the new technology might be more appropriate. Community-wide discussion of frameworks for this decision-making would support the industry.

Post-approval changes (10). There is a strong desire from industry to have a conversation about how to implement changes to processes more efficiently. In general, the FDA was viewed as reasonable in their approach to post-approval changes. There is an aversion to making changes that could have an unanticipated change in patients and a disruption in brand.

Topics that were raised in multiple interviews included:

1. Knowing when analytical data is "good enough" vs. needing clinical data.

2. Ways to accelerate the review cycle timeline.

3. Using knowledge from past experiences in current conversations.

4. Disconnect between company risk assessment and Agency risk assessment.

5. Operating in a global regulatory environment, with the potential need to maintain multiple processes and supply chains to meet different Health Authority requirements.

6. Desire for greater reliance on internal quality systems to manage post-approval changes, especially minor changes.

7. Differentiating between magnitude of process change and magnitude of product change.

8. Suggestion that the potential for more interaction throughout the post-approval process could allow organizations more insight into the process and could enable the ability to respond to questions or gather additional data on an accelerated timeline if necessary.

3.2.3. Operating in a Global Regulatory Environment (11).

All companies operate in a global regulatory environment and every company that participated reflected that this is a huge challenge. Each health authority has different timeline requirements, requires different discussions, and requires different amounts and types of information. There was recognition that this is not an FDA-specific nor CDER-specific problem.

One common sentiment was that it is not a practical business decision to run and maintain a supply for two different processes for different groups of countries, and this has a significant impact on whether or not to pursue process changes.

There is an appreciation that the FDA engages with the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH), the World Health Organization (WHO), and a number of different international forums. Although the burden of harmonization is not on the FDA, companies were hopeful that the FDA has the opportunity to influence and take a leadership role in the global regulatory environment. There is a desire for alignment on the science and the risks associated with the science. If the top markets were aligned in their interpretation of ICH guidance and in their requirements, that would alleviate a huge burden. Specific markets mentioned were the United States, the European Union, Japan, Canada, Brazil, and China. In addition, the WHO could be leveraged to reach alignment across a larger number of countries with their own internal mechanisms.

3.2.4. Consistency across the FDA (8).

There is a desire for consistency across the FDA, which includes between individuals (e.g. different reviewers or inspectors), between the review and the inspection process, and between leadership (“the podium”) and reviewers/inspectors.

Between individuals (reviewers or inspectors) (7). There was acknowledgment that each person is an individual, but there were notable differences between seasoned and newer members of the Agency staff. Experience and comfort level could drive these differences and could, in part, be addressed by education and training. An additional suggestion was an internal Agency quality control process that, for example, pairs up reviewers for training.

Between leadership/policy and reviewers/inspectors (3). It appears that the Agency as a whole, and particularly leadership and policymakers, want to move technology forward and embrace innovative approaches and novel strategies. However, the implementation of these approaches at the review and inspection level is often considerably more conservative than the messages received from the podium. The perception is that there is less comfort with new technologies and approaches, and the associated real or perceived risks, at the individual level. Education and training, increased early interaction, and perhaps other mechanisms could be developed to improve comfort levels of individual reviewers, in alignment with the goals set out by leadership in support of technology advancement and innovation.

The disparity between leadership and day-to-day processes can also be seen at scientific conferences. Leadership from the Agency often encourage people to engage in new technologies as appropriate, but the reviewers tend to take a conservative approach to filings. Encouraging and improving scientific sharing and collaboration could improve consistency between messages heard in public forums and experienced in day-to-day practice.

Between reviewers and inspectors (3). There was a sense that the line between review and inspection may be blurred, even though the objectives of these functions are different. It was acknowledged that inspectors are generally diligent about knowing where questions cross the line from what is directly in front of them to something that needs to go back to the reviewer. It was reflected that having a product reviewer on site can result in a different (and more positive) interaction and there was encouragement for CDER to maintain this practice. Additional education and training are also potential mechanisms to improve consistency and understanding of boundaries by all parties.

3.2.5. Guidance Documents (8).

Companies are constantly seeking updated and new guidance documents from the Agency. A revisit of “Changes to an Approved Application for Specified Biotechnology and Specified Synthetic Biological Products” (2) would be welcome. It was also reflected that the principles in guidance documents might still apply, but the examples provided might be out of date (specifically mentioned was ICH Q5E). It was reflected that the industry would benefit from faster release of guidance, especially related new technologies and approaches, acknowledging that there is a process to issue such guidance. There was also a desire for less formal input from the agency in the form of case studies, best practices, or other learnings that could be shared to identify gaps in submissions or successful approaches to improve the quality of submissions overall.

ICH Q12 (3). The ICH has developed a draft guidance to harmonize post-approval CMC changes, ICH Q12 (1). As of December 2019, the guideline entitled “Technical and Regulatory Considerations for Pharmaceutical Product Lifecycle Management” had been adopted (Step 4 of the formal adoption procedure) and is ready for regulatory implementation in ICH regions. Companies are hopeful that the new ICH Q12 guidelines will shape the future regulatory environment. The document has the potential, if successful, to deliver on streamlining the review process for commercial products, which is estimated to currently take 3–5 years.

There is concern, however, about how the different regulatory authorities might interpret ICH Q12, related to operating in a global regulatory environment. This could be a great opportunity for the FDA to take a leadership role in alignment on the concepts outlined in ICH Q12.

There was a suggestion that a public–private partnership could facilitate a way to work through the details of what this guidance might look like in practice through a case study.

3.2.6. Manufacturing Process Development (7).

There were several discussions around barriers to implementing new technologies centered around manufacturing process development.

Specification setting (4). Process specifications are currently based on small numbers of clinical lots and represent a relatively homogenous data set, whereas a pool of manufacturing data can be quite different as molecules are often manufactured at multiple facilities. Companies desire operational and scale-up flexibility to match any needed lot size. It would be helpful for industry and the Agency to collaboratively examine commonly measured attributes in terms of what they do and do not represent for the safety and efficacy of the product.

One potential idea was that a product could gain preliminary approval with an initial (and small) number of clinical batches, and then come back at a later date with more manufacturing data to set appropriate specifications. Another idea centered around using prior knowledge in regulatory filings to set specifications broader than those set by clinical experience alone.

Using prior knowledge in regulatory filings (6). A wealth of platform-based data, for both processes and molecules, exist within industry. Use of platform data to support regulatory filings could expedite and simplify the regulatory submission and review process. There were questions around how information used in one application could be cross referenced, if similar and applicable, to another product at a different site, to leverage prior knowledge without putting in an independent submission. A broader scientific conversation could be held around what science, technology, or prior art can be deployed to widen specifications.

3.2.7. Additional Regulatory Pathways/Tools (7).

There were a number of ideas for additional regulatory pathways or tools to lower the business risk and accelerate the timeline for implementing emerging technologies in biopharmaceutical manufacturing (with the goal of accelerating patient access to medicines while still assuring safety, efficacy, and supply).

1. Drug master file equivalent for new technologies—This type of tool could be used for technologies referenced by multiple organizations or used by the same organization for a multi-site or a multi-product submission (e.g., platform technologies; see Using prior knowledge in regulatory filings).

2. Breakthrough/accelerated/priority/fast-track technology designation—This tool would provide an expedited pathway, ideally at the post-approval stage. Such a tool could provide incentive within a company and reduce the timeline risk for using a new technology in a program with high value.

3. More interaction during submissions and approvals—Particularly with respect to post-approval changes, more engagement throughout the process could enable a company to respond more quickly.

4. Vendor interaction—A public–private partnership, like NIIMBL, could be used to enable direct vendor engagement with the Agency to gain scientific understanding of new technologies, assays, and approaches.

3.2.8. Education and Training (5).

It was reflected that both industry and the FDA require a highly trained staff. The industry as a whole could benefit from more applied education, such as bioinformatics or applied process engineering. The ability to adopt new technologies (from both sponsors as well as regulators) relies on scientific understanding of new technologies and better education drives a "comfort level'. Additionally, all parties could benefit from better education on risk-based approaches.

4.1.1. Interaction between the Industry and the FDA.

On the topic of interaction between the industry and the FDA, there was a recognition that there is a large diversity in approaches both within the FDA and within a given company. There was a general feeling that the ability to talk to a reviewer has decreased over the past decade, and that CDER reviewers are more formal than reviewers from other centers. There was a desire to be able to have a conversation based around data, which may or may not be complete, toward understanding if a particular approach might be promising. There was broad agreement among industry representatives that fear of receiving a final, formal “no” from the FDA can be a barrier. For the culture between industry and the FDA to change and to improve the nature of informal interactions, there should be a reduction in fear that everything the FDA says will be taken as formal guidance, and the industry must "stop taking everything" the FDA says as formal guidance.

4.1.2. Changes to Approved Manufacturing Processes.

There was significant discussion around post-approval changes to manufacturing processes, and particularly the business case around these changes. There was discussion about the different risks to implementing new technologies at different stages, including risks to timelines and the complexities of managing changes in a global environment. There was discussion about how best to bridge old and new technologies. There was a consensus that it is important to be driven by science; however, technologies with good science may still experience a delay in approval because of unfamiliarity. There was a desire to better understand successful approaches taken by other organizations but a simultaneous understanding of the need to protect a competitive edge or proprietary information. Industry representatives shared that there have been examples of new technologies being used to successfully widen specifications, but this requires a risk on the part of the company. There is a significant widespread aversion, resulting in internal barriers, to the possibility of needing clinical data to implement a change to a manufacturing process.

4.1.3. Consistency across the FDA.

A final significant area of discussion was around the context and consistency of reviewer questions. Representatives suggested that there could be opportunities for training or exchange of knowledge. Coupled with the desire for more interaction, there may be a way for the Agency to evaluate and prioritize efforts that promote additional feedback in the areas where they want to see companies invest and grow.

4.2.1. Business Case for Adoption of New Technologies.

The afternoon discussion began around the components of the business case for the adoption of new technologies. It was agreed that to say there is “no” business case is an oversimplification. There may be a business case, but it is perceived as not strong enough to overcome hurdles, whether real or perceived. There are separate challenges for implementing new technologies in early-stage projects, which have a lower likelihood of making it to market, as well as in late-stage projects where timelines are especially important and which also have the added challenges of a global regulatory environment. The nature of the global regulatory environment also limits opportunities to implement new technologies post approval. From the company perspective, there is a desire to implement a new technology for multiple products at the same time. Companies are constantly seeking confidence that the path to approval with a new technology will not be more difficult or require significantly more time than using a more well-known technology. Time delays may be unacceptable to certain programs, and the uncertainty of the timeline of the path to regulatory approval for processes using new technologies is often enough to keep a company from implementing the new technology.

There was discussion on the differences between cost and value, as well as risk and opportunity. There was also discussion on whether the cultural drivers for an overall conservative approach are because of significant past high-impact events or the “sum of fears.”

There was agreement that Contract Manufacturing Organizations (CMOs) will have an even harder time adopting new technologies until the industry reaches a “consensus” because of the need to convince multiple companies to buy their manufacturing capacity.

4.2.2. Interaction between Industry and the FDA.

There was much discussion about the types of interactions, and the industry representatives' desire for more informal interaction. Informal feedback could help organizations internally de-risk implementing a new technology; however, culture changes are required to be able to have these types of interactions in a productive way. It was agreed that there appears to be a direct relationship between the specificity of a question and the formality of the response, with more specific questions requiring more formal interactions. Further, operating within complex organizations is a challenge for both the industry and the Agency, as different individuals and groups interface at different points in the process.

A safer space for interaction could center around scientific considerations such as risks, issues to be resolved, and parameters to be considered, and less about specific issues. Workshops can be a useful venue to have precompetitive discussions, but every organization has a different definition of when a conversation strays into competitive territory. A public comparability database could be another venue for precompetitive sharing of scientific data. The community could also benefit from more discussion of failures as well as successes.

Case studies. There was significant discussion around case studies as a tool to facilitate interaction and to establish a common vocabulary. A-Mab (3) was a significant effort and was widely viewed as a positive experience, but was published in 2008. There are other ongoing efforts such as A-Vax (4) and A-Gene (A-Gene is an ongoing effort supported by the Alliance for Regenerative Medicine and NIIMBL). There was enthusiasm from both the industry and the Agency for potential future case studies as resources allow.

Collaboration and shared learnings. There was agreement between the industry representatives and the Agency that shared learnings can make everyone more effective in their roles; however, there was no good suggestion as to how those might be collected and shared. The FDA does not “own the information” and individual organizations do not share regulatory successes at open meetings on a regular basis.

3.2.3. Changes to Approved Manufacturing Processes.

There was discussion around the comparability protocol as a tool for post-approval process changes. One suggestion was a workshop around a novel and theoretical manufacturing product or process to identify what the risks are and are not, in the context of implementing the technology in a post-approval context.

3.2.4. Consistency across the FDA.

There was a significant discussion around consistency of reviewer questions during information requests. There is a desire from industry for more context about why a particular question might be asked. It was reflected in multiple interviews that there were experiences of getting no questions for a given technology in one process, and many questions around that same technology in a similar process, which creates risk. It’s important to know when there are significant differences in lines of questioning and for all parties to understand when a certain context is important. There are inefficiencies on both sides, with information requests that don't clearly articulate the reason for the question, and not understanding the question and not fully responding, which can lead to many rounds of written back-and-forth that can add significant delays. These delays can be internal to companies as well. There was a discussion about the value of identifying the missing information, providing a reason for why it is needed, relating the reason to safety and/or efficacy, and making the specific request.

Every company and every reviewer have some experience that drives how they approach different situations and therefore there are challenges around individuals with many years of experience as well as challenges with more junior staff. This is an opportunity where NIIMBL could support the community by brokering the collection and sharing of relevant and appropriate experiences.

When discussing new regulatory tools, a key consideration is the benefit to public health. For example, if implementing a new manufacturing technology can help relieve a drug shortage, there is a clearer benefit to public health, and this could generate additional conversations. There are other mechanisms, such as legislation, that could help incentivize innovation and support repatriation of biopharmaceutical manufacturing to lower the risk to the quality, consistency, and availability of biopharmaceuticals.