An edited version of this article appeared in the August 2020 issue of BioPharm International.
OR WAIT 15 SECS
Cynthia A. Challener, PhD, is a contributing editor to BioPharm International.
New therapeutics modalities and the need for greater process efficiency are driving technology development.
Chromatographic separation is a fundamental unit operation in the downstream processing of biologics due to its ability to separate desired biomolecules from a diverse array of impurities. As the range of therapeutic modalities has expanded, the need for newer solutions has driven the development of new approaches to chromatographic separations and new resin chemistries and column designs. Similarly, the desire for more efficient and cost-effective, large-scale chromatography systems is leading to advances in single-use solutions and systems for continuous, multi-column chromatography, as well as the development of real-time monitoring capabilities. Despite being a traditional technology, chromatography continues to evolve in order to meet the latest needs in biologics manufacturing.
Although widely implemented, chromatographic science is generally complex and requires an in-depth understanding to maximize its benefits, according to Hemanth Kaligotla, manager of application specialists for Sartorius North America’s Chromatography Task Force. In addition, he notes that extensive process development is needed to fine-tune the operational parameters (conductivity, pH, salt type, buffer system) and a correct packing procedure to ensure reproducibility at different scales. Furthermore, because chromatography can represent a significant portion of the cost of goods in downstream processing (DSP) operations, proper optimization of capacity and selectivity is required.
An edited version of this article appeared in the August 2020 issue of BioPharm International.
For most biological drug products, one of the major challenges is the separation of product-related impurities, according to Alejandro Becerra, senior field applications scientist at Thermo Fisher Scientific, including the separation of homo- and hetero-dimers in bispecifics, charge variants in monoclonal antibodies (mAbs), full and empty capsids in viral vectors, and supercoiled plasmid from linear and open circular isoforms and double stranded RNA from messenger RNA constructs. “These closely related species have very similar physicochemical characteristics, but ion exchange, hydrophobic interaction, or mixed-mode chromatography can still be used to address these challenges, although more optimization may be required,” he observes.
Adding to the complexity is the fact that each specific chromatography process varies depending on the specific biomolecule to be purified, how it is produced, and the relative amounts of the target product and the impurities, says Becerra. “For example, a mAb that is a secreted product and is present at a relatively high concentration poses different challenges compared to a viral vector such as an adeno-associated virus (AAV) that is intracellularly produced requiring cell lysis and its concentration is orders of magnitude lower than that of a mAb,” he explains.
Even where platform purification processes exist, it is necessary to identify whether the products fit on the platform and pursue purification development accordingly, notes Jean Aucamp, team lead for R&D at Lonza Biologics in Slough, United Kingdom. Kaligotla agrees. “Each protein is unique, with different stability properties, isoelectric points, molecular weights, etc. Thus, for each new protein process, a number of screening exercises have to be performed to design the DSP train, which often involves the use of three to four orthogonal chemistry steps to achieve the desired level of purity.” For newer platforms, adds Aucamp, it is also essential to build in robustness and broad coverage. “Diverse products require a modular development approach tailored to the product attributes,” he concludes.
New molecules with varying upstream titers pose additional challenges to the development of efficient and cost-effective downstream purification, according to Mark Chavez, senior principal scientist with Fujifilm Diosynth Biotechnologies. “Yield is often more critical for low-titer processes, whereas volume and throughput may be limiting for high-titer processes,” he comments. One approach Chavez suggests is to establish an equipment-based platform, such as three chromatography steps and one tangential flow filtration operations, which allows for purification technique flexibility while maintaining facility fit.
While templated purification processes for standard mAbs are well established with many purpose-built tools available, Darren Verlenden, head of bioprocessing at MilliporeSigma, points out that mAb-like molecules (bi-specific antibodies, FC-fusion proteins, etc.) and novel modalities (viral vectors, plasmid DNA, cell therapies) are relatively new to the market and many are lacking effective chromatographic tools and templates.
“The physical and chemical differences of mAb-like molecules tend to negatively affect the performance of existing chromatographic tools,” Verlenden notes. Novel modalities are typically much larger in size and have different binding sites compared with traditional mAbs. A large molecule, such as viruses (>100 nm), plasmid, or even a large protein complex, has a lower diffusion and mass transfer rate, according to Kaligotla. “For such products, we need to use chromatography solutions based on convective flow rather than diffusion flow, as a result, leading to the use of membrane chromatography for the purification of viral vectors and plasmid DNA,” he says.
“In addition to the different properties of novel therapies, the lack of templated purification processes presents a challenge to end users with respect to achieving their yield and purity targets within a reasonable timeframe with available resources,” Verlenden says.
The rapid growth of the market for novel modalities is therefore putting a strain on many process development organizations to develop processes without the right tools. “Templates,” Verlenden asserts, “would help reduce the time and resources needed to develop a manufacturing process.” Becerra adds that for novel biomolecules, like viral vectors, the capture step is a real challenge, and the ability to have an affinity capture step that could be set up as a platform similar to Protein A with mAbs is needed.
Challenges to chromatographic purification also differ depending on the maturity and stage of the biologic molecule to be purified, according to Henrik Ihre, director of custom-designed media, Cytiva. “Processes for approved biologics are more centered around productivity, process economy, robustness, and security of supply of key raw materials and other essential products to support manufacturing. For novel drug targets, in earlier clinical phases, the key challenges may rather be centered around high-throughput purification and analysis based on generic platforms to reduce time to market and afford a predictable manufacturing process during scale up and validation,” he explains.
Over the past two decades, Ihre believes that a minor revolution in resin technologies has occurred. Capacity and throughput optimization in the early 2000s were followed by the introduction of multimodal ligands designed for specific, large-scale applications and advances in the chemical stability and productivity of Protein A resins for mAb purification.
More recently, he notes that as new (antibody fragments, mAb conjugates, bi-specifics, etc.) and some older (pDNA, oligonucleotides, etc.) modalities started to emerge in clinical pipelines, steps have also been taken to develop new resins offering high throughput and high productivity solutions for the next generations of biologics. Some are based on a novel high productivity base matrix combined with an already existing ligand and others on novel and engineered ligands addressing new customer needs for selectivity and stability. “In general,” he observes, “working closely with customers in setting the critical quality attributes has been a more successful strategy in recent years than the R&D-driven approach common a decade or so ago.”
As one specific example, Chioma Chimezie, manager, downstream process and manufacturing sciences at Immunomedics, points to the introduction of Protein A resins that allow the use of low concentrations of sodium hydroxide and have improved success rates in manufacturing facilities. “Cleaning and sanitization are usual risks with chromatography and especially with processes containing Protein A, which is obtained from cultured or engineered Staphylococcus aureus or Escherichia coli. These risks have been substantially lowered with resins that can be cleaned with sodium hydroxide,” she says.
The availability of mixed mode resins that have robust ion exchange performance while offering a wider operational window and increased resolution on the resin separation functionality have also been important for improving purifications at scale, according to Chimezie. “Having dual and unique functional groups (e.g., hydrophobic and ionic) in combination with high throughput design-of-experiment screening procedures provides more opportunities to develop robust processing and design space,” agrees Chandrika Ediriwickrema, principal scientist/group leader with Fujifilm Diosynth Biotechnologies. Products making use of composite materials may also enable the combination of separation steps for streamlined purification processes, Aucamp notes.
Another trend, according to Kaligotla, is to apply a tailored affinity for new processes. “The experience and familiarity with Protein A in the antibody sector (and also [immobilized metal affinity chromatography purification] IMAC purification at laboratory scale) certainly drive the desire to have an effective and simple capture step that results in high purity,” he comments. These new affinity chromatography resins can be integrated into newer platforms to broaden coverage or upgrade existing platforms, adds Aucamp.
In the case of affinity capture for AAV, Becerra points to the development of a pan-affinity resin by Thermo Fisher Scientific that is able to bind all different AAV serotypes, including engineered capsids, and has enabled end users to move towards platform processes. He notes that the same ligand technology has also been used for the development of resins specific for the affinity capture of antibody fragments.
Ion exchange resins have also become more fine-tuned, allowing more selective binding of either target molecules or impurities, according to Verlenden. One of MilliporeSigma’s recent developments is a cation exchange resin designed to enable preferential binding of unwanted impurities such as protein aggregates, allowing for the target molecule to flow through the column. “This chromatographic mode offers end users significant advantages compared to the typical ‘bind/elute’ approach, including intensified processing, reduced costs, and enhanced ease of use. End users experience significant reduction in resin and buffer volumes, driving smaller manufacturing footprints for subsequent downstream processes, including smaller columns, buffer tanks, virus filtration setups, and ultrafiltration equipment,” he asserts.
In addition to the resin chemistry, the pore size, bead size, and rigidity of chromatography resins are also critical and, in some cases, have been optimized to address challenges specific to large biomolecules like viral vectors and nucleic acids, Becerra observes.
With respect to base bead chemistry, customers have a couple of choices, ranging from natural polymers with hydrophilic properties that show low nonspecific adsorption (e.g., agarose and cellulose) to synthetic polymers with different levels of hydrophobicity, according to Kaligotla. “These chemistries are used to generate particles of different sizes that are suitable for different targeted applications in preparative biochromatography,” he says.
While many new resins have particle sizes in the 40–50-micron range, it is important to note, says Becerra, that the permeability of a packed bed also depends on the rigidity of the resin backbone. “As such, columns packed with materials like polystyrene divinylbenzene result in much lower pressure drops compared to highly cross-linked agarose or polymethacrylate chromatography media of the same particle diameter,” he explains.
Given the significant and ongoing advances in resin chemistry, one of the key differentiators going forward for resin vendors will be simplification, according to Kaligotla. “Winners in the race are the vendors that support process simplification (going from a three-step workflow to a two-step workflow), process intensification (availability in prepacked formats, ready to apply), and process integration (simple connection to systems, flow paths, etc.). Customers are mainly challenged by the effort they need to put into screening experiments. Vendors that support their customers in this direction with screening tools (prepacked screening devices and recommended protocols as well as software tools for analysis) give them an additional advantage. This becomes especially important for mixed-mode functionalities, where property/function relationships are not clearly understood,” he explains.
Higher binding capacities are also desired, according to both Chimezie and Aucamp, both for traditional columns and membrane absorbers. “Increased capacity of resins allows for fewer chromatography cycles and therefore fewer column change-outs or repacks. In addition, Protein A chromatography typically has elevated resin costs and limited lifetimes when compared to subsequent ion exchange-based chromatography resins. Therefore, further focus on resin binding capacity and sanitization/storage to elongate use is an opportunity under active evaluation,” Chimezie remarks.
Resin backbones continue to evolve, providing better bead support with increasing ligand densities to improve capacity, adds Chavez. Ligand libraries and in-silico modeling are also being used to design custom affinity resins to purify specific products and/or hard-to-purify products, speeding up the development process, he comments.
Meanwhile, robust affinity ligands using mild elution conditions are one of the hot topics right now because many new molecules in the pipeline are more sensitive compared to traditional mAbs, according to Verlenden. The reduction of chromatographic steps is one additional goal for this kind of material.
In addition to advances in both base matrix and ligand technologies, novel formats and solutions such as fibro-based devices are entering the market and offering a more open pore structure where mass transfer is governed by convective flow, according to Avril Vermunt, strategic technology partnerships leader at Cytiva. “These structures allow high protein binding capacities at very short residence times by using rapid cycling chromatography, resulting in cycle times of minutes instead of the hours needed for resin-based chromatography,” she says.
These newer solutions are needed for new modalities such as viral vectors, conjugated mAbs, and plasmids, which as large molecules sometimes cannot diffuse into the porous structure of a typical chromatography bead designed for traditional proteins, resulting in slow kinetics and low binding capacities, according to Ihre. “New formats such as fibers, instead of porous beads, may offer not only high throughput and short cycle times, but also a higher accessibility for binding throughout the porous structure and device allowing for improved productivity,” he comments.
Indeed, membrane chromatography configurations are becoming more prevalent with more functional groups becoming available, according to Ediriwickrema. “Membrane chromatography operated in flow-through mode offers tremendous throughput with potential continuous integration with adjacent process steps,” she states.
Membrane absorbers, monoliths, and fibers have their own challenges, however, according to Becerra, including the limited sizes available for scale up and process development. Therefore, a more rational design of pore size, bead size, and ligand density specifically suited for these targets is needed, he says. For capture chromatography, Becerra also points out that there is still a need for affinity resins specific to molecules like lentivirus with ligands that allow elution of these molecules under conditions that keep the virus particle intact and functional. Similarly, most resins for other chromatography modes were developed for the purification of proteins and not for larger molecules.
Becerra also highlights newer technologies such as cassette devices with scaffolds that support chromatography beads and provide additional wall support, enabling the use of high flow rates with soft or semi-rigid resins. There has also been some resurgence of radial flow columns for novel modalities that generally have low product concentrations and thus require high volume loadings, he says. “This type of format effectively functions as a short axial bed, permitting the use of higher flow rates with acceptable pressure drops,” he explains.
In addition to new technologies, Becerra notes that the use of alternative approaches to chromatographic separations such as hybrid pH/conductivity gradients and displacement chromatography have also proven beneficial to address some of the current purification challenges of biologics. He also observes that the fundamental understanding and implementation of other adsorbents like activated carbon filters in the purification of biologics has also increased in recent years. “This technology has proven successful in the removal of process-related impurities like host-cell proteins and small product-related impurities such as fragments,” he says.
Two key advances have been achieved in recent years with respect to column design—easier-to-pack systems and the availability of pre-packed options—and both help minimize operational risk, according to Chimezie. “Easier-to-pack columns ensure a more reliable and consistent distribution of resin, while both can, depending on the operation and diversity of products being purified, boost facility run times,” she asserts.
Pre-packed columns are increasingly used at the pilot scale, where risk is shifted to the column supplier, adds Kaligotla. As they become more readily available and in larger sizes, they are also alleviating time and resources in the manufacturing suite, according to Ediriwickrema.
Column designs that allow for automated operations, including column packing, are further reducing risk. “This type of technology has allowed more reproducible packing results and consequently reduced the time and labor spent in column-packing activities,” Becerra observes. “The introduction of automation has improved the repeatability of pack integrity results with minimal user intervention. The prescribed recipes from the supplier can further assist in packing various media and resin classes from multiple suppliers. As a result, column packing has moved away from being an art and into a science, where the variables of column packing can be more easily fixed,” adds Kaligotla.
Closed processing is another important trend impacting chromatography operations, according to Verlenden. “The market is calling for specific tools that allow for bioprocessing in a completely closed manner, in which the process components are segregated from the external environment and employees. A good example of this need is with larger molecules such as vaccines, where no final filtration is possible at the end of the process due to their inability to pass through a sterile filter. Therefore, the chromatography process must be sterile from the start. All of the components of the column (hardware and resin) or membrane device need to be sterilizable with the ability to maintain performance after exposure to sterilization methods such as gamma irradiation for sterilization,” he explains.
The increased need for flexibility, greater prevalence of smaller lot sizes, and high diversity of different drug targets are driving the need for smaller pre-packed and single-use chromatography columns. Shorter bed heights are also requested to enable shorter residence times, according to Jakob Liderfelt, global product manager for resin technologies at Cytiva. “Smaller diameter columns can easily facilitate short bed heights, while it is a challenge when the diameter is increased,” he says.
In addition, for some biological products, the mass needed is much smaller compared to mAbs, but the capacity of the resins currently used in these applications is relatively high, according to Becerra. “There is a need for columns with diameters smaller than current offerings in pre-packed and manually packed formats with sanitary designs such that they can be used in CGMP [current good manufacturing practice] processes,” he remarks.
As the number of novel and more personalized therapies addressing single individuals or smaller patient populations reach the market, there will also be a need for smaller-scale operations and the development of efficient column formats, Liderfelt adds. “Such columns will require proper documentation and design for GMP-regulated processes and in some cases also solutions for sterile or aseptic processing,” he observes.
Despite the rapid evolution of chromatography technology, further developments are needed to boost efficiency and productivity. “Advances in column design that are still needed include highly efficient packing, good flow distribution even at large diameter, inert materials, and minimum pressure drop. These features would enable faster processing while maintaining the quality purification process,” Verlenden asserts.
Chavez also sees several opportunities for further improvements, such as more effective high-throughput small-scale devices to make the translation to large scale more seamless, the ability to repack prepacked columns to better leverage valuable resin resources, additional bed support within large-diameter columns for faster flow rates, and embedded process analytical technology (PAT) systems with columns for real-time analysis and development of new control mechanisms.
Additive manufacturing (3D printing), meanwhile, would enable on-demand preparation of the adsorptive column interior and the column hardware in just one step, according to Verlenden. “While improvements have been made in these areas, there is still a long way to go before these design advancements are fully established,” he notes.
Advances in resin chemistries and column designs have occurred simultaneously with some developments in process control technologies. Systems are available today, for instance, that can measure higher product concentrations inline and eliminate handling intermediate sample offline preparation steps such as dilution, according to Aucamp. He also notes that automated sampling technologies for at-line analysis are also in use today. “[Process analytical technology] PAT advances allow for quick, informed process decisions at a higher quality level, while artificial intelligence maximizes the current information contained in traditional chromatography monitoring (pH, conductivity, UV, pressure) for better process control,” adds Ediriwickrema.
Better use of inline process analytics across the sequence of chromatography steps would provide a significant advance in process controls, which in turn would enhance product quality, according to Chimezie. “We have been exploring better integration of our sensors to determine timing for process transitions such as cleaning, sanitization, and equilibration, eliminating the excess buffer quantities built into the process from the early development stage,” she comments.
Going forward, real-time (<10 sec) aggregate monitoring for cation exchange bind and elute (CEX-B/E) purifications for real-time/automated event control (pooling) has been identified as one of the top 10 attributes for automation in the May 2020 BioPhorum Operations Group in-line monitoring/real-time release testing publication, according to Kaligotla (1). “Significant yield improvements could be made based on real-time aggregate breakthrough because current [ultraviolet] UV-based collection has a high safety margin that cannot be measured in the elution,” Kaligotla says. A high safety margin results in high batch variation in the starting material, which leads to high output variation, he adds. Online charge profiles during CEX B/E and for the DNA and host cell protein content of anion exchange flow-through processes were both identified as needed to improve the potential for real-time release and continuous processing, he also notes.
Real-time process monitoring and control are also the final elements needed to achieve fully closed and integrated processes, according to Verlenden. “The ideal monitoring technologies are rapid, inline, and provide information on multiple process parameters simultaneously. Moving and automating process control decisions to the manufacturing floor also makes continuous processing more viable, Chavez remarks.
Rapid on-line analytics for critical quality attributes are also needed for biosafety testing, Verlenden says. “Most of today’s analytics rely on taking multiple samples and using offline instruments requiring process pauses and processing delays. While many of the current analytical methods may remain the same, we predict advances in automated sampling, speed of testing, and automation of data analysis,” he observes. Recent advances in molecular testing methods for rapid and sensitive viral detection are also bringing the industry closer to the goal of real-time release.
From the user experience perspective, there is also plenty of room for improvement, according to Pete Genest, strategic technology partnerships leader at Cytiva. “Automation interface screens haven’t changed drastically in the past 15 years, and graphics are still very two-dimensional and based on piping and instrumentation diagrams. Use of the interface often requires extensive training from the vendor and lengthy manuals,” he explains.
Cytiva is making an effort to keep the end user in mind when developing new automation platforms and design user interfaces that are as intuitive as possible. Because single-use systems are simpler than traditional stainless-steel equipment with significantly less valving, the human-machine interface screens for these systems are simpler by default, often with less screen navigation, allowing for incorporation of additional information previously not possible to include, he adds.
Smaller and more efficient biomanufacturing facilities prefer single-use technology, according to Kaligotla. “Single-use technology provides increased manufacturing and tech-transfer agility at lower capital and operating costs. Functionally closed single-use manifolds for chromatography skids, prepacked columns, sterile fluid, and filter transfer sets, and robust sensor components also enable long-term processing in a bioburden-reduced environment,” he says.
Other areas of concern include the risk of leaks (flow path integrity over extended use), limited leachable and extractable data, managing particles, potential bioburden issues, and the need to rely on a large network of extended vendors. “The industry is responding to these challenges by developing new single-use films that offer greater strength and resistance to leaks; by introducing more sensitive integrity testing; and by advancing standards, best practices, and transparency around supply-chain, sterility, particles, and leachables,” Verlenden asserts. Kaligotla believes that process intensification, closed systems, novel robust films, PAT, data analytics, and automation will drive future implementation.
Cost is also an issue. “Single-use columns have many advantages and serve as a means to drive operational productivity by reducing or removing non-value-added time associated with column set-up and cleaning/sterilization/validation. However, given the cost of the resin, our analysis always comes down to maximizing the number of cycles versus the column cost,” observes Chimezie. “In multiproduct facilities, it is difficult to have sufficient areas available to store and trend multiple columns if they are not being used to their maximum number of uses,” she adds.
Even so, Becerra indicates that chromatography systems with single-use flow paths are prevalent in multi-product facilities and particularly for clinical programs to avoid cleaning validation. The size and the pump capacity of these systems also fits well with the production of biologics that require small doses, he notes. Ediriwickrema finds that single-use technology is actually more cost-effective for initial implementation and smaller campaigns than stainless steel, and when pre-packed columns are also used, they can be treated as consumables, eliminating activities such as packing and maintenance, as well as the need for storage space.
Smaller systems and columns that are CGMP-suitable are needed for some molecules like viral vectors, though, according to Becerra. “In these applications, the use of reverse flow for elution or cleaning steps can be beneficial and most of the disposable flow path chromatography systems do not offer that option,” he says.
Process intensification is one of the main drivers for reducing manufacturing cost at commercial scale, according to notes Aucamp. In chromatography, that generally means operating continuously using multiple, smaller columns.
The benefits can be significant. Continuous/multicolumn chromatography can greatly reduce the amount of resin required and can better utilize the buffers needed, according to Chimezie. “Since loading is near the column maximum and material is collected more frequently, a more efficient process flow is achieved, leading to less time in the manufacturing suite. Smaller columns with reduced bed heights can be linked in series to cycle product several times to process a comparable volume to current batch processes. The overall result is greatly increased resin utilization and process throughput, and these efficiency improvements can be further enhanced with the use of prepacked columns,” she explains.
Process intensification through continuous chromatography also eliminates hold steps, reduces the equipment footprint, and better accommodates flexible manufacturing, according to Kaligotla. In addition, functionally closed single-use manifolds consisting of robust sensor components enable long-term processing in a bioburden-reduced environment at clinical or commercial scale.
“While these systems are more complex, an increasing number of technical solutions are being commercialized that reduce the implementation challenge,” Aucamp says. For instance, the system design, operating scheme, and automation are much more complex for multi-column chromatography, and innovators and early adopters have had to develop expertise to implement this technology, according to Vermunt. “In order to be more widely adopted, intuitive data visualization and automation are being developed and system design is also being considered to ensure it is fit-for-purpose for each application,” she notes.
Reliability is also crucial for complex multi-column chromatography systems that operate continuously for extended periods. “The ability to detect and prevent adverse events is essential for robust process control. PAT, artificial intelligence, and in-silico modeling are tools that can add to the needed process monitoring capability,” comments Chavez. Verlenden agrees that a key challenge is validating a continuous process to prove process repeatability and reproducibility. “Advances in process control software platforms, automation, and process analytics should help simplify the management of the process and facilitate the validation of such processes,” he says.
Multi-column chromatography technologies have initially been targeting the Protein A capture step, primarily due to high costs of the resins, according to Verlenden, head of bioprocessing at MilliporeSigma. Leaders in the field have demonstrated its use in bench and pilot scale operations and more recently there have been examples of its implementation in a clinical manufacturing setting, Becerra adds.
Meanwhile, the development of all-flow-through processes for the CEX and AEX chromatography steps is an area of opportunity, according to Verlenden. “End users would benefit from the process efficiencies gained through higher productivities and lower costs while meeting their purification targets,” he asserts.
1. BioPhorum Operations Group, “In-line Monitoring/Real-Time Release Testing in Biopharmaceutical Processes—Prioritization and Cost-Benefit Analysis,” www.biophorum.com, May 2020.
Cynthia A. Challener, PhD, is a contributing editor to BioPharm International.
Vol. 33, No. 8
When referring to this article, please cite it as C. Challener, “Chromatographic Science Clarifies Separation Challenges," BioPharm International, 33 (8) 2020.