Communication and taking the time to develop the process are key to successful transfer and scale up of biologics.
By Susan Haigney
Editor's note: The following is an extended version of the article published in the print version of BioPharm International's March 2017 issue:
Tech transfer and scale up of biologics can be as complicated as biologic therapies themselves. During tech transfer, methods, skills, technologies, and importantly, knowledge is transferred from the drug developer to another site or group, often a contract manufacturing organization (CMO). It’s a process that experts say requires time, organization, attention to detail, and communication (1). And developing a well-defined process is crucial: “A structured transfer process is critical to knowing when specific activities should occur and what the expected outcome is from any given stage,” according to Beall and Rushing (1).
According to Mike Ultee, principal of Ulteemit BioConsulting, companies often fail to provide complete information about their biopharmaceutical or protein. He suggests that companies should allow enough time for tech transfer and that they should transfer the process first. Ultee also stresses communication and scientist-to-scientist interaction. “Scientists from similar departments at both the transferring company and the receiving company need to get acquainted, understand the transfer process, and then work side by side at the bench or in the plant,” Ultee told BioPharm International in an Editors’ Series Q&A (2).
To learn more about the challenges of the scale up and tech transfer of biologics, BioPharm International spoke with experts at Lonza; James Blackwell, president of The Windshire Group, LLC; and John Moscariello, vice-president, process development, and Jeremy Young, director, process transfer, at CMC Biologics Seattle.
The challenges of scale up and tech transfer of biologics
BioPharm: What are the challenges when scaling up from lab scale to commercial scale in biologics?
Moscariello and Young (CMC Biologics): Depth filtration and centrifugation are not always scalable for manufacturing operations. However, having both technologies is desirable based on the number of lots and time available to transition to another modality. Significant increases in cell mass have led to limitations with both technologies. Next-generation harvest technologies, including flocculation with cell settling and microfiltration, are being implemented, and the data are limited regarding the utility of their small-scale models.
Blackwell (Windshire): As a chemical engineer, I think first about mass transfer and heat transfer limitations. Fortunately, some of these can be anticipated and planned for in advance. This requires an understanding of what the rate limiting or controlling phenomena are and at each scale. What is limiting or rate controlling at the small scale often is not at the large scale. That is because the time constant for one phenomenon can change differently from another with scale. While mixing, heat transfer and oxygen transfer all become more challenging upon scale up, one of these may become controlling with scale. For example, when single-use bioreactors were a nascent technology to the industry, there were reports that some bags melted on the larger bioreactors. Someone didn’t anticipate the heating and cooling demands at scale, probably because it wasn’t a problem with the small bioreactors. The ratio of volume to surface area decreases with increasing scale. Thus, more heat or cooling needs to occur, relatively speaking, per unit surface area with increasing scale. This is particularly true for microbial fermentations. In the case of microbial fermentation, heating and mass transfer demands can often be controlled by the fed-batch feed rate. Thus, you want to use a feed profile at the bench or pilot scale that will translate to the production scale.
Lonza: When scaling up mammalian processes from lab scale to commercial, the decline phase of the production culture can be faster at larger scale, leading to a premature reduction in product accumulation. Another challenge is a high amount of dissolved CO2 in culture, if not managed carefully, can lead to high osmolality and cessation of cell growth or productivity. Additionally, non-physiological dissolved CO2 levels in the culture can impact the cell’s ability to correctly perform post-translational modification of expressed products.
A typical example of a challenge for scaling up microbial processes would be mis-incorporations or modifications of amino acid in the product by upscaling a fermentation. Other examples are unexpected low titer compared to process development runs or unexpected filter clogging in depth-filtration steps due to different quality of filtration starting material. An example includes a filtration step that is not scalable due to facility restrictions (filter holders/pumps, etc.). An example of scaling up from lab to facility includes defining the ideal stirring conditions, as the stirrer setup in the lab is typically completely different than the mobile tanks that are used in the plant. The stirring in the facility setup must not have a negative impact the on product quality.
[In regards to chromatography steps], sometimes the linear flow needs to be lowered due to the volumetric flow restrictions in the large-scale facility. This would have to be tested during process optimization prior to scaling up.
BioPharm: What challenges and/or considerations must be addressed when outsourcing tech transfer and scale up?
Moscariello and Young (CMC Biologics): In process development, it is important to consider and communicate outlier data. Interpreting a good run as a reproducible run can be a costly mistake, so it is important to include as much data from previous development runs and/or at-scale batches when outsourcing projects. Many companies have experience in development and production, but not every molecule is the same, and the devil is in the details.
Another consideration would be to understand the facility and the resources that are allotted to each project. Does the company have the capacity, the equipment, and the capital for the project? If not, is it committed to having a plan in place to meet a client’s timeline while meeting quality and cost expectations? Substantial investments are usually devoted to technology transfer projects, and the project should feel like a partnership between the parties—both companies should feel equally vested in and accountable for [a project’s] success.
Lonza: A mammalian process designed with insufficient appreciation of operability within a cGMP facility and lack of process understanding from the process transferring partner can increase the risk of uncertainty during process transfer and scale up. However, these issues are not insurmountable if a good collaborative relationship is fostered and a common understanding of the underlying issues are developed.
[In regards to microbial processes], a typical example of a scale-up issue would be a process that has to be transferred that was already produced in a smaller scale. However, we sometimes face the situation that we are confronted with unit operations that are not scalable further (e.g., size-exclusion chromatography). Another typical example [is a] centrifugation step with a beaker centrifuge [that] has to be transferred to a disc stack centrifugation setup.
Blackwell (Windshire): First, are the parties ready? In other words, are they available or have capacity during the timeframe you need? Are they willing to participate? Meaning, will they feel ownership and take responsibility? And very importantly for biopharmaceuticals, are they capable and competent? The latter can best be assessed by reviewing the credentials of those involved and their prior experience and success with tech transfers. If consultants are used, you may want to consider using one organization as opposed to lots of independent consultants. Having lots of independent consultants can turn into a cat herding exercise and [can] be a barrier to knowledge and information sharing and transfer.
BioPharm: Are there specific challenges for the tech transfer of antibody-drug conjugates (ADCs)? Monoclonal antibodies? Cell therapies?
Moscariello and Young (CMC Biologics): The major challenges in technology transfer for any modality exist when there are known gaps in scale-down models or if the scalability from the small-scale to large-scale is unknown; for example, in the scale-up of cell-culture production bioreactors. It is not uncommon to see slight differences not only in how both processes perform, but also in product quality when scaling up from small-scale to pilot scale, to clinical-scale, and then to commercial-scale.
A good example for ADCs involves a mix-and-match approach to the differences that exist between scales. There are many recent advances that have addressed these issues, such as single-use bioreactors (SUBs), which are available at a variety of scales and have shown excellent scalability. However, the largest format that is frequently used is 2000L, which is sometimes not sufficient for commercial production.
CMC Biologics has an approach where we run six 2000L SUBs in parallel, then harvest and combine the product streams downstream to produce a single drug substance lot, with an equivalent amount of product to a commercial-scale 12000L stainless-steel bioreactor. Additionally, this approach significantly minimizes the technology transfer risks as the process is scaled out using multiple 2000L SUBs rather than scaled up to a new bioreactor.
Blackwell (Windshire): I’d like to address cell therapies since that is an area we have been doing more work in recently and [is] a growing area. Many cell therapies are personalized medicines and these represent unique challenges from a manufacturing and supply chain standpoint. One of the things that can aid greatly in tech transfer for these types of processes is thinking far in advance about your manufacturing process, facility, and supply chain requirements. Think simplicity, suitability for purpose, and scalability. Many of these types of processes will not scale by volume or by unit operation, but rather in parallel. This means modular thinking and standardization can ease tech transfers. An important aspect of many of these processes is the human element, which is often a larger consideration than it is even for traditional biologicals, such as antibody manufacturing. This makes clear and good procedures, and training and communications, vital to any tech transfer. In judging the success of transfers, having historical data compiled from the process using a continued process verification solution that pulls quality control and batch record data will enable easy comparisons of the receiving site with the sending site, and operator-to-operator performance. These data should be tied to patient outcomes.
Lonza: The unique characteristics of an antibody-drug conjugate--a small-molecule cytotoxic moiety attached by chemical linker to an antibody--demand unique manufacturing and analytical infrastructure. This infrastructure must provide a working environment that isolates the manufacturing and lab personnel from cytotoxic chemicals while providing the aseptic environment required for manufacturing a biological drug substance.
The small-molecule cytotoxins used to manufacture ADCs are biologically active at the nano-gram level; therefore, residual unconjugated cytotoxin must be removed from the drug substance and product contact equipment following manufacturing. Free cytotoxin removal at the end of the drug substance conjugation process is further complicated by reaction stoichiometry, which often requires cytotoxin in excess of that consumed in the conjugation reaction. Removal of residual cytotoxin is critical for preservation of drug substance safety and is a unique process step relative to the typical biological manufacturing process.
Detection of residual cytotoxin at the PPB level on the surface of manufacturing equipment requires expertise in the development and execution of ELISA or tandem mass spectrometry (MS/MS)-based analytical methods. Depending on the process requirements and the ability to chemically inactivate the cytotoxin, dedicated or single-use manufacturing equipment can eliminate the possibility of cross-contamination.
Ensuring quality and performing validation
BioPharm: What quality procedures specific to biologics must be put in place for technology transfer?
Lonza: Technology transfer processes are critically important for biologics manufacturing, as the demonstration of comparability of biologics made at different phases of the drug-development cycle, at different scales and at different sites, must be assured. Due to the complex nature of biologics and uncertainty of how the biochemical and in-vitro assays truly simulate the mode of action in vivo, exerting control on the manufacturing process is an integral and important facet to maintain consistency and comparability of biologics made initially for product development and subsequently to meet ongoing clinical supply.
The technology transfer procedures implemented are typically guided by International Council for Harmonization (ICH) Q10 or equivalent guidance from other regulatory authorities. This guidance is practically translated into establishing global and local standard operating procedures instructing the implementation of a risk-managed assessment of transferred information from donor to recipient site related to the manufacturing process, analytical methods, and other assays to verify consistency and comparability of products made at the recipient site.
Blackwell (Windshire): Lifecycle document and knowledge management is crucial due to the complexity of biologics. Procedures that govern technical documentation requirements gated to each stage of development ensure that critical information is captured and available for tech transfer. For example, required clinical manufacturing campaign reports will summarize the process used, operating parameter performance, issues and investigations, and other key information. This becomes invaluable to others later, and helps support claims for process characterization and robustness; selection of control and critical parameters; and setting of specifications. A summary of lessons learned should also be included. Templates, formats, and requirements for technical reports, including signatories, ensure the appropriate data, information, and knowledge are captured and properly reviewed. Another important step is a procedure for tech transfer describing responsibilities, activities, and deliverables. This helps in planning and resource allocation and acts as a valuable road map when in the heat of the tech transfer.
Moscariello and Young (CMC Biologics): Quality procedures are going to be specific to a facility and the capability of that facility, which will be documented. For example, classifying the facility as a commercial-phase facility or as an early-phase one depends on the level of quality documents that exist to control the materials in the manufacturing process.
A company should have a quality policy document in place that establishes the minimum requirements prior to manufacturing any biologics. An example of a policy would be to segregate the manufacture of microbial products from manufacture of mammalian products to prevent cross-contamination. Another example would be to define materials that must be free of animal-derived products ahead of time, to prevent sources of viral contamination. Other procedures should also be established to define the local health and safety requirements as well as the company’s internal commitments. An example of that would be to identify hazardous chemicals such as methotrexate, or the disposal of organic compounds.
BioPharm: Which considerations specific to biologics must be made concerning process validation in technology transfer?
Moscariello and Young (CMC Biologics): The major differences in the technology transfer and process validation approach between therapeutic proteins and other small-molecule therapeutics are due to the complex nature of proteins. For small molecules, analytical testing can verify the exact structure of the molecule and impurities, whereas this is not possible for proteins due to post-translational modifications (glycosylation, oxidation, deamidation), charge variants, and potential other conformers. Therefore, there is enhanced scrutiny of the process for making a biologic product and a philosophy that the process defines the product. This philosophy emphasizes the need for a significant amount of process characterization and for a robust validation package. In technology transfer runs for process performance qualification (PPQ)—a concept that has replaced traditional process validation—each parameter should have a criticality classification with a justified normal operating range (NOR) and proven acceptable range (PAR). These classifications and ranges should be supported by process-specific data, including relevant literature or subject matter expertise.
Recent advances in analytical tools—like process analytical technology (PAT) and advances in protein characterization—have led many to question whether the process-defines-the-product approach can be replaced with a product-defines-the-process approach. Regardless, it is important that analytical tools be flexible for use on the manufacturing floor, or in a quality setting for product release. If there is a need for rigorous criticality designation, NOR and PARs, then it is important to have the right tools to adjust the process to provide the desired product. Generally, if there is a need for rigorous criticality designation, a NOR would be appropriate; or if we have the right tools to adjust the process to provide the desired product, a PAR would be used.
Blackwell (Windshire): In terms of traditional validation (“x number of successive lots”), tech transfer and scale-up activities will provide invaluable data that support process understanding, and knowledge obtained can be useful in further developing the control strategy. This sets the stage for successful process qualification. Due to the inherent variability and complexity of biologics, the design space and strategy should take into account future scale-up considerations and risks. This can be met with qualified scale-down models or modeling. Successful scale-up that matches predicted behavior demonstrates good process understanding. This can help in justifying fewer as opposed to more process qualification runs to demonstrate process robustness.
Lonza: Due to the complex nature of protein synthesis within eukaryotic cells and their in-vivo mode of action, reliance on biochemical characterization alone to show product comparability is insufficient. Therefore, the process controls that were identified and demonstrated during process validation studies to maintain consistent and reproducible manufacturing processes and product critical quality attributes (CQAs) within pre-defined limits need to be maintained during the technology transfer to the recipient sites.
A supplementary process risk identification/assessment using ‘cause and effect’ and/or failure mode and effect analysis tools within the recipient site would confirm the capability of the transferred manufacturing process to be maintained under similar control regimes as at the donor site. If this is not confirmed, then supplemental process characterization studies, in addition to the process performance qualification (PPQ) runs, can be proposed for the supplemental market authorization application (sMAA) (e.g., supplemental biologic license application for FDA authorization), for the recipient site(s). However, if it’s already known that the manufacturing process shall undergo technology transfers to different manufacturing facilities, this can be included into the original process risk identification/assessment and the process validation program and thereby avoid, or minimize, the need to perform supplemental process characterization studies after transferring the process into the recipient site(s).
The technology transfer of a biologics manufacturing process with an existing process validation package in place is no different to transfer [than that] of a biologics manufacturing process without an existing process validation package. The process validation data together with other process data from the donor site can be assessed for risk: If substantive gaps exist with operating the transferred process in the recipient facility, remedial evaluation experiments or facility modifications can be proposed to address these gaps, and the sMAA program will also reflect the additional studies needed. Additionally, establishment of scale-down model(s) of the transferred process in the recipient site(s), as well as product-specific analytical methods, will be required to support the product comparability. Other in-vivo tests, if required to support product comparability, are typically proposed and managed by the contract manufacturing organization (CMO) clients.
BioPharm: How can risk in scale up and tech transfer be minimized?
Blackwell (Windshire): It starts with good process understanding and characterization. If risk assessments have been performed on the process, these are a good starting point for understanding where the risks lie during tech transfer. However, the tech transfer process itself presents new and unique risks. Thus, a risk assessment of the technical transfer process itself can be done, starting with gap assessments of those things that will change or are different. Intel Corporation learned a long time ago to minimize any changes during a tech transfer. They took this so far as to replicate in the receiving site a piping u-bend around a support beam that existed at the sending site but not at the receiving site. While it is a noble goal to always strive for no changes, the practical reality is that it is often not entirely achievable for biopharmaceutical processes. Risks associated with the business processes should also be assessed and addressed. Often communications will be identified as an area of risk. This is why many organizations will send a representative(s) from the sending site to the receiving site well before the main thrust of activity.
One particular area that needs special attention is any transfer of analytical methods. This should never be handled as an afterthought. Since the success of the transfer will be monitored largely by these results, anything that brings questions as to the validity of these results will sow the seeds for confusion and time delays during tech transfer. Therefore, time should be allowed for successful transfer of these methods prior to their need.
Another key, often neglected aspect of tech transfer is involvement of the manufacturing [department] early in the development cycle. Ultimately, manufacturing will be the process owner and is the customer of the development function. Early involvement of manufacturing or their technical representative (e.g., technical operations or manufacturing sciences) will help ensure the process is fit for purpose and able to transfer into the receiving facility.
Lonza: The risks in scale up and technology transfer of a manufacturing process can be minimized through the following concerted approaches:
• The manufacturing process should be robustly designed for the intended long-term manufacturing purposes.
• A clear product comparability strategy with deep understanding of process- and product-critical quality attributes and how these influence mode of action in vivo and what manufacturing process inputs impact these parameters is required.
• Structured technology-transfer procedures to assimilate detailed information between transferring sites should be established.
• Close collaborative communication between technology-transfer stakeholders with aligned project goals is very important.
• Good scale-up/scale-down models must be established within recipient sites to demonstrate successful technology transfers between sites and subsequent support of the manufacturing process.
• The manufacturing facilities within a company’s network should be designed and operated according to a common design basis and with harmonized operational practices.
Moscariello and Young (CMC Biologics): A facility and equipment gap analysis is recommended. It is also recommended that you allow time to procure the right equipment and allow engineering runs to utilize the equipment at scale. Manufacturing success typically increases with the number of batches, and there is a steeper curve in the first few batches. It is highly recommended to add at least an N of 1 prior to the first cGMP batch to increase that level of success. Whether the process has new equipment, or a unique feed strategy or collection criteria, providing an opportunity for the manufacturing and transfer teams to see the development at scale under non-cGMP conditions will allow [them] the flexibility to make the appropriate adjustments for future GMP lots.
1. R. Beall and W. Rushing, “Effective Technology Transfer Strategies for Biologic-Drug Development,” Editors’ Series, On Demand Webcast, Executive Summary, BioPharm International.com, January 2017.
2. M. Ultee, “Working Together: Tech Transfer for CDMOs and Biologics Drug Developers,” Editors’ Series, BioPharmInternational.com, Jan. 4, 2017.