Managing Uncertainty in Continuous Biomanufacturing

Published on: 
BioPharm International, BioPharm International-05-01-2018, Volume 31, Issue 5
Pages: 12–17

More published data and initial regulatory approvals are needed to drive adoption of continuous bio-manufacturing.

There is general agreement in the biopharmaceutical industry that new manufacturing paradigms will be essential to tackling issues around the cost and timeliness of new drug development, particularly for complex biologic therapeutics. Process intensification is being pursued as a first step toward fully continuous manufacturing. A combination of business challenges, technology gaps, and regulatory uncertainties must first be overcome before continuous biomanufacturing (CB) becomes a commercial reality. Equipment suppliers, drug manufacturers, and regulators are working closely to address these issues.

The potential benefits to implementing continuous processing range from economic to time savings to quality improvement. Process intensification, the forerunner of continuous manufacturing, is allowing drug manufacturers to achieve, in some cases, up to 75% reduction in cost of goods sold (COGS), according to Andrew Bulpin, head of process solutions for MilliporeSigma.

Business challenges

There are a range of business hurdles, however; benefits vary with each bioprocess and require individual assessment. The historic standardization on fed-batch processes is one key challenge, according to Parrish Galliher, chief technology officer for upstream with GE Healthcare Life Sciences.

In addition, he notes that the business risk for CB is substantially higher due to the 5-10-fold increased rate of raw material consumption under continuous upstream processing (perfusion-medium consumption), which presents a significantly higher financial risk in the event of a product-quality failure compared to batch or fed-batch processes. Conducting quality-by-design (QbD) studies and validation of critical parameters and process design spaces is also more complex, labor-intensive, and time-consuming due to the extended duration of continuous production campaigns. The cost of training can be higher as well; operators need to be able to run more complex processes 24–7 without mishap.

Proof of scalability across upstream and downstream processing is another main hurdle, according to Peter Levison, executive director of business development at Pall Biotech. “Processes must satisfy both internal and external expectations from concept through to commercialization,” he says.

“There are many unsubstantiated claims in the literature that the costs for CB are lower compared to current methods. On the other hand, users are beginning to report that perfusion media consumption is far higher for CB in the upstream, and in the downstream, buffer and resin reductions need to be balanced toward any increased capital or consumables costs compared to batch chromatography,” Galliher says. On the positive side, he does note that CB has the potential to improve product quality consistency in cases where variability in product quality is not acceptable.

Levison stresses, too, that, “While investment may first seem like a financial hurdle, one of the most often overlooked points is the exceptional savings on the backend (even with higher costs on the front end), for the life of the process. Savings come from less human/operator intervention and investment, less cost in process critical resources like fluids, and fewer quality-control demands, which combined lead to improvements in productivity and process economics.”

Duncan Low, previously with Amgen and currently principal at Claymore Biopharm, asserts that from a purely business perspective, there are in fact more incentives than hurdles, “unless you start stacking up uncertainty.” “You have to define the technical issues and identify the risks as for any project. You may ask if you want to start with a commercial molecule or a biosimilar, or a new molecule where the risk of clinical failure overshadows the whole assessment. You may also want to ask about regulatory acceptance, which you could argue is a business risk,” he observes.

Technology (data) gaps

Much of the technology for CB is in a relatively advanced state (compared to that for cell and gene therapies, for example), and much of it has been implemented to varying degrees, Low argues. “CB is, however, broader in scope than continuous processing for small molecules in that it considers the entire process, involves biology as well as chemistry and physics, and can be performed in plastics rather than steel,” he comments.

While suppliers have developed and demonstrated scalable platform technologies for end-to-end CB, there still remains some work to be completed with analytics and automation and control to secure the greatest benefits of improved product quality and safety, according to Levison. “The right level of automation and control has been a hurdle for drug manufacturers, regardless of process type, and is even more crucial for continuous processes because they are run in closed environments. There needs to be sensors and detectors that measure critical process parameters and critical quality attributes in real time, ideally in-line or at-line,” he comments.

“Improvements are needed around sensor and process analytical technologies (PAT), as this would bring enhanced process monitoring and real-time release of the product,” agrees Galliher. Such sensors must be stable to operate in-line for weeks without drift or fouling.

Advances in these technologies are also important in the drive to create and archive more data that prove continuous processing advantages. “Data are really the top technical gap,” Levison emphasizes. “There still remains a paucity of published experimental data and case studies to demonstrate that continuous processes work effectively and provide multiple advantages. Consequently, there has yet to be a strong enough record of commercialized drugs (small or large molecule) to model success from,” he states.

Levison anticipates that the Internet of Things will play a significant role in how the industry will be looking to manage and drive progress based on data. Pall Biotech has also created a process development services lab to help customers overcome the perceived challenges in transitioning from batch to continuous processes, or to create new processes from scratch. “With this approach, it is possible for drug manufacturers to start to gather data with less investment up front,” he says.

Indeed, for Low, the transparent and seamless integration of data between suppliers and manufacturers is desirable when it comes to addressing PAT and automation requirements for CB. Overall, he sees CB as offering “amazing opportunities” for biopharmaceutical manufacturing. “Analytics and process control is an area with significant opportunities to apply ground-breaking, disruptive experience from other industries in data gathering, manipulation and understanding, and machine learning. We are at the beginning of Industry 4.0, and industry groups are working to address the ramifications of the impact these developments will have on the pharma industry,” he explains.



Regulatory uncertainties


Regulatory authorities support the implementation of CB, and companies can work closely with different agencies from the process development stage. Limited regulatory precedent does exist for continuous upstream processing given that more than 15 biologic products manufactured using continuous upstream processing are on the market, according to Galliher. “There is still regulatory uncertainty in the downstream space, however, because no biologics produced continuously in the downstream [process] have yet been licensed for commercial manufacturing,” he says.

In particular, regulatory uncertainty surrounding batch definitions for continuous processes remains. Bulpin points to two examples: demonstrating product equivalence for a perfusion process at day one through day 60 and ensuring virus validation in an integrated and continuous bioprocess. “In the former case, the cell-culture development team must express product with low variability and the purification development team must minimize the impact of variability in the cell-culture feed on the final profile of the purified product. Even if successful, it is not possible to produce a fully uniform product over the entire process time, so we must be able to define a level of functional equivalence that is acceptable to the regulators,” he says.

For virus validation, which is a downstream operation, Bulpin says it is necessary to rethink the entire strategy for CB because for integrated and continuous processing, variability in the performance of one operation can have a ripple effect on several downstream steps. “This impact as well as changes in operating modes including mass loading, flow rates, and consumables re-use, must all be accounted for during virus inactivation and removal validation studies,” Bulpin observes.

There are also questions about how to handle the traditional three batches during a performance qualification when working with 30-90 days processes, according to Galliher.

“Until these types of questions are addressed, manufacturers have to consider performing QbD studies over the full extent of a continuous bioprocess campaign, driving up material costs, time for process development, and labor requirements,” Galliher says.

Based on the process development and validation strategy for a product, a batch may be defined according to the mass of product produced, such as 2 kg of purified product equals one batch; or process time, for example, 24 hours of clarified cell culture harvest from a perfusion bioreactor; or as a number of process cycles typically in the downstream process, that is 50 cycles through a Protein A column on a multi-column capture step.

“While regulators appear agnostic on which definition is used, determining the required data package to support the definition and successfully collect that data in a way that is transparent and defensible will be the critical step in timely approval of the process for commercial production,” Bulpin asserts. “Suppliers, biomanufacturers, and regulators must work together to build the regulatory framework for approval of next-generation processes,” he adds.

Integrated difficulties

The ultimate goal with CB is to perform integrated biomanufacturing, with upstream and downstream operations connected together so the process is running in continuous mode. Doing so brings additional complexities and hurdles, according to Galliher.

First, liquid flow rates through all unit operations must be matched and controlled continuously to be in synchrony, so that no overflows or flow stoppages occur. The process controls for each of the unit operations must also be synchronized to maintain steady-state operations without mishap, which requires sensors to operate stably and reliably over extended periods of time.

“It is increasingly important with CB to define process and control strategies and use in/at-line analysis and a step-wise approach to ensure control over each individual process step before verifying the complete process. Setting the design-space for integrated CB often starts with defining the design-space for one-unit operation at a time, followed by multivariate data analysis to define the complete process,” Galliher notes. He adds that reliable model-scale setups should be available, including those for the connections between unit operations, where process development and trouble-shooting can be performed.

Sensors and automation controls are relevant to bioburden control as well, which is also essential throughout the duration of integrated continuous processes. Connection of the bioreactor with the clarification system presents particular challenges, according to Levison. “In fed-batch, there is less chance for bioburden contamination due to the relatively short time taken to harvest the cell culture (typically 2–8 hours). In the case of perfusion culture, however, where the user may be continuing this process for 20–60 days or even longer, there must be a closed system to maintain low bioburden; consequently the analytics, sensors, and controls become so much more important. The user needs to be able to ‘know’ what is going on during the entire process,” he states.

Today, the industry is taking an evolutionary approach toward achieving integrated and continuous processes. “Process intensification, like multi-column chromatography or the implementation of single-pass tangential-flow filtration, has been a successful first step. We are now in the process of connecting operations like anion exchange and virus filtration into a single operation,” Bulpin says. “As we move toward connected and continuous processes, software must enable processing systems to communicate and integrate several operations in order to maintain process controls and product quality. Biomanufacturers are clamoring for this type of software and control platform,” he continues. “Based on the knowledge we are gaining in the earlier stages of this evolutionary path, we will innovate and develop software and automation capabilities that don’t yet exist to fully connect and integrate both upstream and downstream processes to enable continuous processing,” Bulpin asserts.

Low agrees: “We must have a realistic appreciation of the issues that need to be addressed and try to tackle them. That will involve identifying the issues, defining them (regulatory, quality, technical, etc.), recruiting the appropriate subject matter experts, and working together to resolve them.”


Setting standards

Given that continuous bioprocessing generally considers the generation of the active molecule as well as its purification and formulation, there are multiple issues where standards can facilitate implementation, according to Low, who also is an officer for ASTM International’s E55 Committee on Manufacture of Pharmaceutical and Biopharmaceutical Products, and vice-chairman of the Subcommittee E55.04 General Biopharmaceutical Products.

A proposal for a new standard on continuous bioprocessing of biopharmaceuticals has been brought to the attention of the E55.04 Subcommittee and will be a focus of the group during 2018. Particular areas of interest, according to Low, include the stability of cell lines over multiple generations, the management of bioburden over extended periods of time under conditions that are conducive to growth, and the design of unit operations such as bind and elute chromatography and viral inactivation via low pH and/or detergent that require defined hold times. Because single-use technology provides many advantages--and has its own set of issues for which standards development is actively ongoing--its impact on continuous bioprocessing must also be considered.

“Collaborating to develop standards in anticipation of the issues rather than reacting after the fact is going to be helpful to manufacturers, suppliers, regulators, and most importantly to patients,” Low asserts. “If the pharmaceutical community genuinely shares a passion for serving patients, we have to collaborate when we can advance common, non-competitive interests for patients in areas of safety, quality, and technology. Consensus standards are recognized by the US government, FDA, and international bodies as a format for capturing and sharing knowledge. Individual and corporate commitments ensure ongoing dialog and the ultimate quality of the finished product,” he adds.

FDA reaffirms support for continuous processing

In February 2018, FDA Commissioner Scott Gottlieb, MD, announced that the agency’s requested budget is intended to continue to fund its current programs at consistent levels (1). It also includes an additional $400 million to support planned initiatives “aimed at supporting new and ongoing efforts to foster more investment and innovation in the development of therapeutics and diagnostics,” including the movement toward continuous manufacturing as a means “to improve the agility, flexibility, cost, and robustness of manufacturing processes.”

Gottlieb also commented in the statement: “Armed with a robust scientific understanding of the requirements and the impact of these advanced manufacturing technologies, the FDA can help industry make investments in these new technologies and grow these opportunities. By developing a science-based framework that includes the regulatory tools and guidance for how products developed in these systems will be evaluated, and by funding research, development, and testing of the enabling technologies, the agency can help reduce the cost and uncertainty of adopting these new manufacturing platforms. The FDA would lead stakeholders in the development of clear scientific standards, policy, and guidance to support the effective and efficient adoption of these new manufacturing platforms, including the new inspectional methods they’ll require.”

Promising future despite the challenges

There is a perception, according to Levison, that transitioning to continuous processing from batch is difficult, when it really is not. “While early adopters may have to work out some of the nuances, they also get the greatest advantage. Single-use technologies were initially viewed with skepticism but have gained acceptance and are now moving into a more mature phase; late adopters have had to play more catch up. At Pall Biotech, we are driven by the potential of CB and are happy to embrace early adopters and work with them to advance continuous bioprocessing,” he states.

MilliporeSigma is also bullish about the future of CB. “We estimate that next-generation processing cuts down one-quarter of original cost, around $100/g to $25/g, and frees up capacity in the range of 25-93% for manufacturing of other molecules,” says Bulpin. He also notes that PAT offers the opportunity for easier and more cost-effective quality and process controls. “Through continuous and automated monitoring driven by an advanced software platform, we can eliminate the need for sampling and offline measurement in many cases. Instead, we will have the ability to make near real-time decisions and react quickly to control our processes and ensure product quality,” he observes.

As a result, next-generation bioprocessing will have a significant impact on how biopharmaceutical manufacturers bring therapies to market, delivering them to patients faster and more cost-effectively than ever before, according to Bulpin. “We estimate that by 2025, next-generation bioprocessing will represent 20% of pipeline molecule revenue and will be a [$1.5-billion] €1.2-billion opportunity for life-science suppliers,” he concludes.


1. Scott Gottlieb, “Statement from FDA Commissioner Scott Gottlieb, M.D., on Administration’s request for new FDA funding to promote innovation and broaden patient access through competition,” FDA Statement, Feb. 13, 2018.

Article Details

BioPharm International
Vol. 31, No. 5
May 2018
Pages: 12–17


When referring to this article, please cite it as C. Challener, "Managing Uncertainty in Continuous Biomanufacturing" BioPharm International 31 (5) 2018.