As many pharmaceutical companies begin to leverage the principles outlined by the Quality by Design initiatives described by the US Food and Drug Administration in the Guidances for Industry [1-4], the way in which firms generate supporting data, and ensure long term process robustness is rapidly evolving. Developing a process control strategy from an enhanced development approach [1] involves a greater degree of understanding of the impact of a particular synthetic route or formulation process to both material attributes (i.e., raw materials) as well as process parameters in both a univariate as well as multivariate sense. Building enhanced knowledge, assessing risks intrinsic to a process, and presenting information in a clear and logical fashion are vital to demonstrating control and understanding of a process to a regulatory agency. These remain key deliverables and responsibilities for the asset owners.
In parallel with this evolution of strategies for process development internally, the pharmaceutical industry has been trending towards a supply chain that leverages third-party networks for research, development, and manufacturing. This increase in third party reliance may include the entire supply chain up through Active Pharmaceutical Ingredients (APIs) being developed, optimized, and ultimately commercialized as long term manufacturing assets in a vendor’s facilities. As a result of this increased outsourcing, there are also evolving strategies and risks directly tied to the justification and subsequent acceptance of starting material designation [4], resulting in a potential increase in the number of registered steps. As a result, the rigor put into outsourced process development and robustness analysis must be commensurate with the risks to the patient, and result in an outcome equivalent to an in-house developed asset.
More recent guidances, such as ICH Q10 [3], clearly describe the role that Pharmaceutical Quality Systems (PQS) play in managing the control and review of all outsourced activities. The recent FDA draft guidance [5] clarifies the specific expectations for oversight of a contract facility (CFs, Suppliers, Partners) by the innovator pharmaceutical companies, (Owners) who are ultimately responsible for the quality of the materials regardless of their place of manufacture.
While responsibilities for GMP activities are becoming clearer, there are no specific guidances on the management or expectations of the relationship between the Owner and the CF for development activities, including those activities that are discussed in ICH Q8, Q9 and Q11 and those that occur in parallel with the GMP activities themselves. The degree to which pharmaceutical companies are leveraging an external network may vary by company, size, internal infrastructure, business needs, and capability. The following generalizations illustrate several of the common arrangements between the pharmaceutical company and the third party contract facility:
- Primarily in-house laboratory effort for process design (e.g., route selection, process optimization), and in-house facilities for clinical trial (CT) manufacture as well as long-term commercial manufacturing. Situational use of contract facilities, and transfers of processes to a chosen sourcing partner depending on internal capacity requirements and specific processing needs.
- Utilization of a sourcing partner or contract facility as a selected extension of the Owner’s labs to partner in the process design and development based upon internal capacity needs. Leverage contract facilities for clinical trial production and data generation prior to commercialization activities. Use data generated during CT manufacture to facilitate transfer of processes back into internal manufacturing infrastructure.
- Little or no laboratory presence is maintained at the pharmaceutical company. The core expert scientists at a pharmaceutical company lead a team of scientists at the outsourcing partner that carry out all the process development activities. Data generated during the development effort is used to support transfer of processes into contracted facilities for commercial manufacture.
In all cases, including situations not highlighted here, many important aspects remain consistent, independent of the arrangements set up with the contract facility. Issues that specifically relate to process learning and control strategy that commonly need to be addressed may include:
- Communication, cultural differences, regional regulatory and environmental requirements
- Harmonization of documentation, and procedures
- Technology transfer, support, and responsibilities
- Harmonization of expectations
- Scale up and facility-specific requirements
- Data generation requirements
- Maintenance of the knowledge database
- Assessment and management of risk
Additionally, understanding and agreeing upon management of change is a critical element of any outsourcing activity and is a key component of any PQS. Elements of all these listed interactions, including (but not limited to) those defined in in ICHQ7 (GMP for active pharmaceutical ingredients) and ICH Q10, require particular attention. Proposed changes based upon a supplier’s recommendations that result in a change to development objectives or manufacturing processes require consideration of many elements such as:
- Do the changes meet the expected quality (CQAs) for the intended use; clinical or commercial?
- Are impurities or physical characteristics (per ICHQ11) affected in a manner that may impact downstream synthetic work or ultimately drug product processing?
- If the material is produced using a new or modified route or condition, has it been evaluated for the presence of new potential impurities (including an assessment of risk relating to test procedures’ capability to detect any new impurities with regard to the synthetic route)?
The evaluation process may include, but is not limited to, the following analyses:
- Structured risk assessments to evaluate whether the change impacts the control strategy;
- Laboratory experiments and/or scale up studies to confirm the product is of appropriate quality;
- Evaluation of the need to verify the equivalence of drug substance manufactured after the change through performance in drug product;
- Assessment of the potential impact of the change on the product stability and/or its degradation pathways.
Below are three complementary perspectives on the performance expectations for suppliers to be able to support an enhanced submission for an API. They provide progressive considerations: starting with strategic issues of supplier engagement, moving to decision making on quality issues and information requirements during development, and finally to some of the practical issues of incorporating supplier-gained experience into a workable knowledge base.
Situational Users of the Third Party Network
John Lepore, Merck
Timothy Watson, Pfizer
For the purpose of this article, the term “situational users” is intended to describe firms that have a stated intent to leverage third party suppliers and external innovation, but evaluate opportunities on a case-by-case basis. This may apply to companies that have retained internal development capacity. In these cases, the roles played by third parties are multifocal, and decisions about which nodes in a combined internal/external network are leveraged are driven by a number of considerations. Some companies have multiple supply bases, in which early chemistry is developed externally, and then as the processes mature and are optimized (internally or externally), can be shifted to a longer term supply base consistent with routine supply.
Decisions might often be determined by the complexity of the synthesis, the classification of the API, (high potency, antibiotic, hormones, etc.), and risks associated with supply chain complexity. There are many other considerations that relate to the strategy of a third party supplier base; however, some of the more obvious dimensions are illustrated in Figure 1.
Figure 1. Dimensions for consideration when choosing a contract facility for partnering.For a third party to take on a role in developing data supporting an enhanced submission, a number of factors need to be considered. First, there needs to be a mutual understanding of complexity and the resulting process risk. In the case of relatively low risk and/or less complex development efforts, an external partner may be able to take a more traditional approach. For higher risk, more complex development issues, there is value in selecting a partner with significant experience in enhanced development and process design capabilities.
As a matter of principle, if the risk being controlled by a third party is very narrow in scope, such as for a regulatory starting material, it would not be unreasonable to expect the partner to control the input to the sponsor’s process by way of specification. As the scope of the activity by the third party becomes more comprehensive, perhaps processing up to final intermediates, or even drug substance, it is generally expected that a higher level of activity by the third party takes place. This can be in the form of enhanced development at either the sponsor or third party supplier and can focus on improving understanding or control of the process leading up to the intermediate to be transferred between parties, as well as working on the downstream process to render it more fault tolerant and enable control solely by specification.
The paradigm of how process information evolves, whether internal to an innovator company, or external within its supply chain, relies on many factors, largely governed by the dynamics between the firm and the third party. There are many variants present today in the industry. Programs can be developed from a technology perspective, and then replicated at scale externally. Alternatively, the innovator may develop a design space as one element of the control strategy, but entrust the details of operationalizing a process to the third party, i.e., the third party is responsible for the overall control strategy and any risk assessments. Lastly, processes can be developed in a fully leveraged mode, where experimental plans are proposed, and after some engagement with the sponsor, are then executed and operationalized.
Cornerstones for externally leveraged innovation include a robust knowledge-sharing paradigm, and well-documented activities on both sides of the partnership. In addition, appropriate intellectual property controls, and a well-crafted quality agreement are also necessary in order to protect mutual interests, and allow the sponsor to appropriately discharge their responsibilities for the end-to-end supply chain.
Users of the Third Party Network for Development and CT Materials Supply
Kevin Seibert, Eli Lilly
Divesting of internal capability to execute clinical manufacturing activities has led to a signifi cant change in the way in which data generation must occur from early through late-stage process development. Many of the considerations for selecting partners to be used for information generation as well as for CT supply manufacture remains identical to former paradigms when situational use of the contract facilities was more the norm (as above). While much of the laboratory eff ort remains unchanged, potential learning that would have resulted from a pilot plant campaign must now be managed at a contract facility. This often leads to a disconnect in expectations which must be managed from the outset of the Request For Proposal (RFP) stage of engagement, all the way through the completion of campaign summary reports.
As Owner companies prepare RFPs, the clear setting of expectations from the outset of conversations with a contract facility will establish not only the material requirements but that of information needs as well. Clear communication of the information needs will allow a contract facility to prepare quotes that refl ect the potentially increased analytical burden as well as equipment and facilities needs to accommodate the information deliverables. Some examples of requested information that may pose a challenge for contract facilities are listed below.
Mass Balance Data
Clear separation of processing streams, obtaining representative samples, and getting accurate weights and volumes of processing and waste streams can pose a signifi cant challenge to facilities attempting to minimize tank usage. Additionally, for those contract facilities lacking appropriate equipment for collecting masses of waste and process streams (drums, appropriate sized scales, fl ow totalizers, etc.), or those that heavily leverage bulk waste containers, separating processing streams and gaining adequate totals may not be possible.
Analytical Burden
In addition to obtaining analytical processing data for product release, inprocess controls and forward-processing decisions, obtaining analytical data for waste and secondary streams may also pose challenges when analytical support is limited. When attempting to clearly articulate a control strategy, including fate and purge studies for impurities, accurate, and timely data are a key part of campaign executions. Understanding the expectations in terms of numbers of samples as well as types of analytical tools needed to perform this analysis is important at the outset of the discussions between Owners and contract facilities. Timely analysis of materials with stability sensitivity, non-standard or expensive analyses (ICP, XRPD, etc.) can also lead to an underestimation of the resources and expenses associated with the analytical component of a campaign executed in a contract facility.
Processing Samples
Additional samples, once easy to obtain and set aside for further analysis and testing, can now present a challenge to an Owner working with a contract facility. Clearly articulating the needs of the Owner for development and subsequent analysis of intermediate processing streams, solids, slurries, and wet-cakes at the outset is also essential. Ensuring equipment is equipped with sampling devices for these types of material must be understood before equipment selection and setups can be performed at a contract facility.
Lastly, as with the situational users of the third party networks, as assets progress toward the latest phases of development, and processes begin to look very much like the intended commercialized process, further emphasis may be placed on generating data for the purposes of a submission, in addition to the production of clinical trial material. Executing campaigns to support the registration stability data package may require several lots executed in a consistent fashion. Alternatively, processes may be intentionally perturbed toward a processing extreme to gain data in support of a proposed design space data set or aid in validating a design space model. Having the fl exibility to execute a campaign against varied processing conditions may pose a challenge to a contract facility, but may be of utmost importance to an Owner company generating data for regulatory purposes.
Users of the Third Party Network for all Development and Commercial Activities
Sue Wollowitz, Medivation
Having all development and scale-up activities carried out by contract laboratories and contract facilities means that managing the supplier relationships, individually and collectively is especially important. The suppliers must truly be extensions of the Owner; quality of capabilities and scientifi c expertise at the facility, as well as transparency, are critical to being able to develop a robust process. The business model also places additional burdens on the Owner to ensure that appropriate data is obtained and that the experiences are collected into a useful knowledge base.
It is often the case that route scouting and manufacture of a pre-clinical safety lot and Phase 1 API are carried out by one contract facility while later manufacture is carried out by a second. In addition, separate contract facilities may be employed for analytical development, polymorph and salt screening, stability studies, etc. Contract facilities that are used for drug product development also gather data, such as criticality of solid state characteristics, that must be integrated into the API control strategy as well. Contract facilities may see themselves as the keepers of information, but it must be realized that in the end, the full knowledge base must be owned by the Owner and the various parts must fi t together. Without that, information gathered early on from one supplier may not be usable to support controlled changes or deviation investigations at a second supplier, or to defi ne design space. While an Owner can maintain a “rosetta stone” to connect data from one contract facility to another after the fact, it is better for the suppliers to gather data and create documentation with an Owner-centric plan and intent from the outset. Below are some of the challenges in developing a supportable and clear knowledge base from a multi-supplier development program.
Harmonization of Procedures
Examples for data expectations to support change control procedures have already been discussed. Even with sampling plans and testing agreed to however, procedures for collecting and documenting the data may vary substantially between the Owner and their contract facilities. In development and campaign reports for example, expectations for depth of content, inclusion of raw data, level of data analysis, and responsibility for drawing conclusions can vary considerably. Selection of acceptable ranges for design of experiments and critical process parameters are sources of variability as well. Even small differences can be problematic when it comes to validation and qualification activities. For example, analytical development strategies and details of validation/qualification protocols vary among Owners and contract facilities and may result in unusable data, repeating activities, multiple documentation re-writes, unexpected costs and delays if these are not made clear up front.
Harmonization of Processes and Equipment
Even within a single facility, equipment may vary in design, but this is certainly the case across unrelated facilities. Unique equipment features may even be a differentiating factor that a supplier offers. Some differences such as in situ monitoring capabilities or drying equipment may be identified as supplier selection criteria since the requirements may be recognized early on. Others however, such as variations in materials addition and transfer procedures, temperature control, sampling methods, and stirrer configurations may be found critical for some processes but only after the fact or when trying to transfer the process among facilities. It is inevitable that these situations will occur, especially if we are to continue to develop innovative manufacturing facilities, but the more a facility knows how their equipment and procedures are equivalent, or not, to industry standards, the more quickly problems may be identified and solutions found.
Harmonization of Nomenclature
Finally, in a network with multiple sources of information, even the nomenclature and definitions require careful attention. Each supplier typically has internal procedures for assigning product codes and/or product identifiers for each material and at the time of submission each raw material, intermediate, API, impurity, etc., may have several identifiers that complicate that ability to trace information across suppliers. Ensuring that Ownerselected nomenclature is also included in development reports, campaign reports, analytical and validation documents, increases the usability of the data. Likewise, definitions and terminology used for materials and processes vary across suppliers; standard terms should be incorporated into critical documents and reports and jargon should be avoided.
Summary
Guidances for Industry ICHQ7, Q8, Q9, Q10 and Q11 have all resulted in the pharmaceutical industry re-thinking their development strategies. Enhanced data packages and information generation are now an even more significant component of the development effort from route selection to the final phases of process optimization. In parallel, the pharmaceutical industry as a whole is finding it necessary to develop relationships and leverage third parties to a far greater extent than what has been traditionally the case and therefore the deliverables from contracted laboratories and manufacturing facilities must support the goals of an enhanced development program. The degree to which a partnering laboratory or contract facilities may be engaged by a specific pharmaceutical company will vary significantly depending on many different criteria. These requests will likely differ in strategic approach depending on the owner company, and in conjunction with the broader scope of engagement with third parties, and will likely result in an even greater need for flexibility from these relationships. Communication, mutual understanding of expectations, and flexibility will be of utmost importance in leveraging activities at contracted facilities for an enhanced submission for drug substances.
References
- ICH Q8 Pharmaceutical Development, (R2); U.S. Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research (CDER): Rockville; 2009.
- ICH Q9 Quality Risk Management; U.S. Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research (CDER): Rockville; 2006.
- ICH Q10 Pharmaceutical Quality System; U.S. Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research (CDER): Rockville; 2009.
- ICH Q11 Pharmaceutical Quality System; U.S. Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research (CDER): Rockville; 2012.
- FDA’s draft guidance for industry; Contract Manufacturing Arrangements for Drugs: Quality Agreements, May 2013.
Author Biographies
John Lepore, Ph.D., Chemical Process Development and Commercialization, Merck Sharp and Dohme Corp., leads API process development for Merck synthetic compounds. John has led the API launch of many products, operation of API pilot plants, sourcing, and compliance functions. John serves on the ISPE PQLI Technical Steering Committee, is member of the Development Technical Committee for the Product Quality Research Institute (PQRI), as well as a number of PhRMA teams. John holds a Ph.D. in Chemical and Biochemical Engineering from Rutgers University.
Timothy Watson, Ph.D., Research Fellow, GCMC Advisory Office at Pfizer, is a member of the PhRMA Expert Working Group (EWG) on the ICHQ11 regulatory guidance document for drug substance. He worked for Marion Merrell Dow in chemical research and development as a process chemist before joining Pfizer, where he continued with process chemistry development responsibilities with a focus on post proof of concept projects. Watson holds a Ph.D. from The Ohio State University.
Kevin D. Seibert, Ph.D., Senior Engineering Advisor, Small Molecule Design and Development, Eli Lilly, is a member of the QbD working group within the IQ consortium, AIChE, and ACS. Kevin has been a contributor to a variety of publications and presentations in the area of Quality by Design and received the AIChE QbD award in 2010 for outstanding contribution to QbD for drug substance. Kevin holds a BS from the University of Colorado, and a MS/PhD from the University of Michigan.
Susan Wollowitz, Ph.D., Vice President of Pharmaceutical Operations, Medivation, Inc., has worked in discovery and process and product development under a variety of operational and collaborative business models. She has worked previously for Dow Chemical, Cerus Corporation, and independently, as well as in collaboration with a number of other companies engaged in partnered, licensed or contract drug substance development activities. Sue holds a Ph.D. in Chemistry from the University of Wisconsin.