Advances in Filtration Technology

Filtration is widely used during biologics manufacturing to remove process impurities and contaminants (particulates, viruses) and for bioburden control. Successful filtration operations are therefore essential to successful biotherapeutic production. As bioprocesses move toward intensification and therapeutic modalities increase in physical size and biochemical complexity, the role of filtration continues to evolve.

Optimization of filtration steps for throughput and product recovery while achieving acceptable process economics is becoming more challenging. Collaborations between filtration technology suppliers, biopharma manufacturers, and regulators are enabling the development and adoption of new solutions to meet these challenges.

Several filtration technologies to manage

Filtration technologies fall into four main categories, according to Mike Collins, director of research and development science for filtration management at Cytiva. They include prefilters, such as depth filters for initial clarification; membrane filters for sterilizing final product, but also for bioburden reduction during secondary clarification or following viral inactivation (VI); virus filters for removal of adventitious and endogenous viruses; and tangential-flow filters for concentration of process fluids, for diafiltration to exchange buffer solutions, and for removal of impurities.

Depth filtration, tangential-flow filtration (TFF), and sterile filtration are commonly used within all processes across numerous modalities, including recombinant proteins, antibody-based therapies, and viral vectors, notes Chris Nieder, director of viral vector downstream process development at SK pharmteco. “The choice of membrane materials and molecular-weight cutoff (MWCO) varies depending on the application, and vendors play a crucial role in providing a diverse portfolio of solutions. In recent years, vendors have also focused on sterilized TFF technology, enabling closed processing when necessary,” he says.

While mainly used for removal of cells and cellular debris and to reduce turbidity and soluble impurities at cell harvest, depth filters are also employed following low-pH VI for monoclonal antibodies (mAbs) to reduce turbidity and impurities prior to polishing chromatography, Nieder observes. TFF, meanwhile, is used in many applications, such as prior to capture chromatography to reduce volume (and thus load time) for viral vectors and buffer exchange for mAbs, for buffer exchange between capture and polishing chromatography, and for concentration and buffer exchange for final formulation.

Most depth filters are non-woven media layers or silica media reinforced with fibers and formed into a sheet, according to Paul Beckett, senior strategic product manager at MilliporeSigma. “The legacy materials for depth filters are cellulose fibers and diatomaceous earth (DE, organic silica), bound together with a positively charged resin to bind nucleic acid contaminants,” he comments. The positive charge helps remove nucleic acid contaminants. The silica, either as DE or a synthetic equivalent, adsorbs larger molecules and particles. Increasingly, activated carbon is used because of its ability to adsorb smaller molecules, Beckett adds.

Membrane filtration in biologic manufacturing is deployed to clear smaller particles, typically below 0.5 microns in size, and can be used for bioburden control, virus retention, concentration, and buffer exchange, states Elliott Zell, global new product marketing–bioprocessing filtration with Solventum. Flat or pleated sheet membranes are used in sterilizing grade (e.g., 0.2-micron) filters for bioburden control in many unit operations and for assuring the sterility of final drug products during fill/finish, he notes.

Membrane filters for aseptic (sterile) filtration typically comprise surface-modified polyether sulfone (PES) and are designed for high performance, both in terms of capacity and flux, according to Beckett. They can also be combined into multi-layer devices for challenging feeds, 0.1-µm variants for mycoplasma protection around the bioreactor, and single-layer variants for high-speed buffer filtration.

TFF filters are generally composed of regenerated cellulose when used for high-value applications (mAbs, viral vectors) and when organic solvents are employed (PES is not compatible), while PES membranes find use in robust, workhorse applications like plasma processing, Beckett observes. Cassette devices are most common due to their easy scalability and because they can be reused, but he notes self-contained capsule variants exist for clinical manufacture campaigns and when containment is important, such as for lentiviral (LV) vectors and antibody-drug conjugate (ADC) manufacturing.

Beckett also notes that TFF is increasingly used in upstream perfusion-based cell culture to remove spent media and/or product. These TFF filters, though, have designs (hollow fiber) that differ significantly from those used in downstream processing and typically are made of modified PES or polyvinylidene difluoride (PVDF).

Vent filters are an unglamorous and often overlooked category of filters, says Beckett. They are installed on tanks to maintain atmospheric pressure and enable blow-down and integrity-testing operations. Many contain polytetrafluoroethylene (PTFE) membranes and are sterilized using steam. PES filters are used in niche applications when sterilization by irradiation is required.

Hollow-fiber membranes can also be in nanofiltration devices for clearance of endogenous and adventitious viruses, according to Zell. “Due to the small size of viruses (typically <100 nanometers), membrane filtration can be highly effective at removing viruses while allowing the drug substance to pass through,” he says.

Nanofilters, agrees Jean Aucamp, associate director of biologics R&D for Lonza, are ubiquitous in mammalian bioprocesses due to their ability to effectively remove both endogenous and adventitious viruses.

Different modalities pose different filtration challenges

Many challenges must be overcome to achieve successful filtration operations, and those challenges vary depending on the modality, the desired goals, and the specific filtration operation.

Monoclonal antibodies (mAbs)

Although manufacturing platforms for mAbs are well-established, challenges to filtration have increased significantly in recent years and in various ways. Improvements in upstream cell culture have, according to Collins, resulted in higher production titers of 5–10 g/liter or more compared to 1–2 g/L a decade ago. “This increase in titer generally comes with higher cell densities (up to 40x106/mL in fed batch), higher turbidity, and an increased level of impurities [host-cell proteins (HCP) and host-cell DNA] that must be removed,” he explains.

While Beckett believes large-scale clarification operations involving continuous centrifugation followed by fine depth filtration can handle these acceptably, two-stage depth filters can struggle to obtain high enough capacities to enable facility fit and economic operation. There is consequently a need for higher capacity clarification filters. Functionalized depth filters are also needed to address the increased level of soluble impurities, which if not removed can impact subsequent chromatographic purification steps, Collins notes.

The higher upstream titers for recombinant proteins are also creating challenges for membrane/aseptic filtration when performed earlier in the downstream process using intermediate membrane filters, according to Beckett. Here again the issue is insufficient capacity. “Specialist intermediate filters that have very high capacity for particulate loads are a viable solution here,” he notes.

There are also challenges during final formulation due to the trend toward higher concentration products (200–300 g/L protein) intended for subcutaneous delivery, according to Nieder. The higher concentrations lead to high viscosity and thus high TFF pressures. “Managing these high pressures in a manufacturing setting can be a concern,” he says. Sterile filtration can also be difficult, with the high viscosities requiring lower flow rates, which can lead to extended processing times. “Oversizing sterile filters can compensate, but then holdup volumes lead to product dilution, which creates new issues,” Nieder comments.

Collins points to high product losses due to nonspecific binding, high non-recoverable volumes, and high pre flush requirements as examples.

Finally, Aucamp observes that nanofiltration for virus removal is challenging mainly due to its high cost and the variation in filter performance seen for different products. In addition, low product concentrations are necessary to achieve sufficiently high flux and short process times, which is against the trend for higher concentrations and leads to larger process volumes. The latter may to some extent negate process-intensification and volume-reduction measures introduced upstream.

On a positive note, Beckett points out that existing filtration technology for viral clearance of standard mAbs is extremely effective. “The smallest parvovirus we are expected to remove to >104 log10 reduction value is not that much bigger than a mAb, and we expect very high product yields (with) short process times for this step,” he comments.

Bispecific antibodies

In addition to the challenges outlined for traditional mAbs, viral filtration, according to Beckett, also presents difficulties for bispecific antibodies. “These molecule types have a tendency to aggregate, which greatly reduces capacity on virus filters, and they are larger than standard mAbs. Achieving economic capacities therefore becomes an issue with these challenging feeds,” says.

ADCs

ADCs are unique in the field of biologic drugs due to their highly potent nature and resultant need to protect operators and the environment from exposure to these complex molecules. “Unlike most bioprocesses, the key challenge here is protecting the operator from the toxic product more than protecting the product itself from the environment,” remarks Beckett. “That can be an issue due to the limited availability of high-quality closed devices that provide appropriate containment during manufacture,” he continues.

Viral vectors

Most depth filters on the market were designed with clarification of positively charged mAbs in mind, which can greatly impact step recovery for viral vectors, according to Nieder. The silica component and positive charge of many depth filters typically bind LV vectors quite strongly, says Beckett, as they are both large and have a net negative charge.

Adeno-associated viral (AAV) vectors, while smaller than LVs, also are negatively charged, and existing depth filters have shown low product yields due to virus adsorption for both types of vectors, Collins adds. One alternative approach is to use lower-capacity filters made of cellulose or synthetic equivalents.

Lentivirus sterile filtration can oftentimes lead to low filtration capacity and low recovery, according to Nieder. That is because, explains Beckett, of their large size and tendency to aggregate, which prevent their easy passage through sterilizing grade membrane filters (0.22µm or smaller).

The other challenging filtration step for viral vectors, particularly LVs, is virus filtration to remove undesired viral contaminants. Indeed, due to the sizes of viral vectors (AAV is approximately 20 nm and LV is approximately 120 nm), it is not possible to use the virus filters typically employed in mAb processes, according to Collins. “There is consequently a higher risk of adventitious virus getting into the product stream, and therefore, more emphasis must be placed on reducing that risk by using methods such as closed single-use technologies and applying virus filtration to any process fluids entering the process, such as cell culture fluids and buffers,” he says.

Using a completely closed sterile process that would preclude the need for sterile filtration creates its own set of challenges, however. Beckett highlights the need for closed and sterilized equipment for TFF in the right pore size, which may not be readily available. Adjusting the formulation to reduce aggregation may help for some viral vectors. He also notes that other large, fragile entities such as exosomes and lipid nanoparticles face similar challenges.

Plasmid DNA

Plasmid DNA (pDNA) products have similar issues to those of viral vectors for clarification depth filtration given that nucleic acids have a strong negative charge. Yields for pDNA can be quite low, and often formulation changes and/or the use of non-charged media is/are required, says Beckett.

Evolutionary advances

While advances are occurring in the chemistries used for membrane media, many technology developments are related to new ways of applying existing chemistries and new device designs, Beckett remarks. “As such, advances must be seen within the wider context of the bioprocessing industry as a whole, which is itself evolving. Manufacturers are under intense pressure to reduce product costs and increase speed to market, which is impacting all aspects of bioprocessing. Therefore, although novel solutions have been introduced, many recent advances in filtration technology have been evolutionary rather than revolutionary, and some novel approaches, such as flocculation for harvest pretreatment, have been borrowed from adjacent industries,” he concludes.

Single-pass TFF (SPTFF) is an example of a new application for an existing technology. “This new mode of operation (not a new filter type) is becoming more widely adopted for final formulation of high-concentration mAbs. It has been common in academia and is now finally being adopted by industry because, when developed properly, SPTFF results in a steady-state process able to achieve high concentrations (200–300 g/L) and lower system pressures when compared to traditional TFF,” says Nieder.

Sustainability is a consideration

When considering potential new filtration technologies, it is essential—as is the case for all aspects of bioprocess development and manufacturing—to evaluate their potential impact on the sustainability of biologics production operations. “Any new product/equipment/technology must be designed with sustainability in mind,” Beckett states. It could be a design that requires less water for flushing, needs fewer packaging materials, has a smaller footprint, or eliminates the use of hydrofluorocarbons for high-performance integrity testing.“New solutions that provide more sustainable alternatives while addressing important filtration challenges will have a higher likelihood of experiencing widespread adoption,” he concludes.

Noteworthy advances

Even though most advances in filtration technology are evolutionary in nature, some noteworthy developments have led to improvements for various modalities.

Depth filtration has received significant attention. A basic improvement noted by Beckett has been the move to synthetic media, which provides comparable performance combined with greater efficiencies, particularly surrounding flush volumes and validation requirements, which in turn lead to more attractive process economics.

To address the challenge of higher cell densities, titers, and impurity levels for mAbs, approaches have included flocculation prior to clarification, acoustic wave separation, and tangential flow depth filtration (TFDF), observes Alexei Voloshin, global head of bioprocess science–bioprocessing filtration with Solventum. He adds, though, that these solutions still have complexities when scaling to large process volumes.

Anion exchange (AEX) fiber chromatography is another solution that, according to Voloshin, has proven to be scalable up to large commercial manufacturing. “This approach leverages a Q-chemistry functionalized fiber bed in a single-use capsule for clarification of harvested cell culture fluid and combines the ease of filter-like flowthrough operation and scalability with the separation precision of chromatography,” he explains.

Functionalized fiber technology can, in fact, perform chromatographic separations across the entire range of process impurities, from whole cells to cell-culture media components, Voloshin contends. Commercial solutions, he adds, enable step changes in the speed of process design, scale-up, and improvements in process economics. “Solventum (formerly 3M Health Care), for instance, offers a functionalized fiber clarification platform designed to accelerate development of mAb candidates expressed in modern, intensified, Chinese hamster ovary cell cultures from discovery to commercial manufacturing,” he says.

TFDF is also becoming more widely adopted for clarification of LV vectors, notes Nieder. “By combining TFF with a membrane having depth like a depth filter, TFDF has led to high recovery and capacity when compared to traditional depth filters,” he observes.

For the larger moieties in the cell and gene therapy space, new separation technologies based on novel ideas are also being explored, according to Voloshin. He highlights functionalized fiber beds and polymer-coated beads offering filter-like flowthrough functionality while trapping process-related contaminants by size, charge, or other characteristics. “In these cases, the line between conventional ideas of filtration and chromatography is blurred to take advantage of both the size and the biochemistry of the drug substance modality,” he notes. “As this field evolves, we’re likely to see much more advanced filter-like process technologies employing advanced physics and chemistries at the core of the separation mechanisms,” Voloshin believes.

In the area of aseptic filtration, basic improvements to device designs are enabling in-situ integrity testing within fill/finish isolators and strengthening of vents on capsules, according to Beckett. More advanced developments involve improved membranes that allow for higher capacity, lower protein binding, and the ability to reduce footprints.

One example of the latter is a new sterile filter technology from Cytiva that address the challenges associated with high-concentration biologic feeds by solving the issue of premature filter blockage from high particle loads. “This new sterile filter can deliver high throughput capacity without the need for oversizing, thus avoiding problems associated with increased hold-up volumes and costly product losses,” Collins says.

Another issue being considered in this area is whether validated, sterilizing-grade filters are needed for intermediate filtrations performed for bioburden/particulate control between unit operations, and usually to protect chromatography columns from fouling and filtration of non-critical buffers in early downstream operations. “Given a substantial part of the filter cost is the testing, validation, and documentation to demonstrate the sterilizing grade, a more open membrane or one with less documentation would likely be both cheaper and have significantly higher capacity for the challenging feeds seen early in biopurification processes,” Beckett states.

Finally, Beckett notes that the recent publication of the long-awaited EudraLex Volume 4 Annex I update on the manufacture of sterile products has laid out regulatory expectations very clearly for membrane filtration. “Certain expectations, particularly those regarding pre-use post-sterilization integrity testing (PUPSIT), have driven innovations and technology advances in membrane filtration to ensure regulatory compliance now and for the future,” he says.

According to Beckett, advances in TFF technology include the open ultrafiltration range of membranes (300–1000 kDa nominal MWCO), for which pore-size control issues during membrane manufacture have been overcome, affording more robust separation solutions; innovations in channel screen technology that allow for processing of high-concentration products; and closed sterilized devices that reduce validation overhead for ad hoc clinical campaigns and allow closed processes for product and/or operator safety.

For viral filtration, advancements have focused on further improving the high capacity and robustness of existing membranes to support dirtier feeds and more challenging products such as bispecific antibodies, Beckett says. He adds that as for other filtration applications, with intensified, integrated, and closed processing becoming a greater need, closed designs that can be irradiated are also in development across the industry.

Collaboration essential to development

As the role of filtration is central to bioprocess design and touches many parts of the process, the next generation of filtration and filter-like solutions will undoubtedly be designed by a collaboration of technology providers, biopharmaceutical manufacturers, and regulators, Voloshin observes. Indeed, advances in filtration for bioprocessing do not occur in a vacuum. Drug manufacturers and regulators provide the requirements for new technologies. Manufacturers also play a big role in developing new approaches to using existing technology.

Technology suppliers, and in some cases academic labs, are largely responsible for development of new solutions. “It is critical for all groups to work collaboratively to progress new technologies effectively,” Collins states.

Manufacturers, adds Beckett, perform the critical role of informing suppliers about their greatest pain points, where the bottlenecks with processing exist, and how a product would need to perform to overcome barriers to entry associated with change control in a regulated process.

“Collaboration with vendors is a great way to assist in getting new products to market,” Nieder agrees. Vendor collaboration can also present opportunities to evaluate products in early-stage development and provide feedback to improve design and performance, he observes.

Changing expectations

The basic idea of filtration has come a significantly long way as the biotechnology industry has evolved. As bioprocess design moves forward, Zell expects to see much more inventive and exciting approaches to filter-like separation technologies.

“The basic premise of completely flow-through biopharmaceutical processes is on the horizon, and it is the advanced filtration solutions of the future that will enable safer, more efficacious, and more accessible biopharmaceutical drugs around the world,” Zell avers.

Beckett adds that there is a paradigm shift underway regarding the manufacturing of biopharmaceuticals. Changes include new pressures on the cost of goods, a shortage of blockbuster drugs that require 60,000-L bioreactor capacity, and new classes of biopharmaceuticals such as ADCs and viral gene therapies that are much more potent than existing classes and therefore require less volume.

“Filtration technology must fit into what the bioprocess facility will look like in 2040, not now. That facility is likely to be much smaller, entirely single-use, multi-product ballroom production and designed for agile manufacture of multiple product types in a single line. To support such facilities, filtration technology must be smaller, cheaper, and/or faster, which are the three basic elements of bioprocess intensification,” Beckett concludes.

About the author

Cynthia A. Challener, PhD, is contributing editor to BioPharm International®.

Article details

BioPharm International®
Vol. 37, No. 8
September 2024
Pages: 12–16

Citation

When referring to this article, please cite it as Challener, C.A. Advances in Filtration Technology. BioPharm International 2024 37 (8).