Send to printer »

Feature Articles : Jul 1, 2010 (Vol. 30, No. 13)

Filtration Evolves to Meet Users' Needs

Adaptation Is the Name of the Game for Large-Scale Operations Upstream and Downstream
  • Angelo DePalma, Ph.D.

High protein titers have been blamed, at least in part, for the mismatches of equipment, time, and facilities that result in downstream bottlenecks in chromatography and filtration operations. Less appreciated are upstream challenges due to an enabler of high-titer processing, namely the use of animal component-free media.

“High protein concentrations upstream require custom culture media that can be difficult to filter, or at best have different filter requirements from serum-supplemented media,” observes Anil Kumar, Ph.D., principal R&D engineer at Pall.

Moreover, high titers’ effect on filtration is multiplied throughout the process train, for example, in the huge quantities of buffer required by larger chromatography columns for high-titer processes, the difficulty filtering ever-more-viscous, valuable process fluids, and new concerns over contamination. “Filter manufacturers must consider upstream processing challenges, as well as develop solutions for downstream processing.”

Cell culture engineers believed that animal component-free media would eliminate the problem of mycoplasma contamination, but hope was illusory, as mycoplasma contaminate plant-derived raw materials and water, as well as animal-derived ingredients. In response, Pall developed a triple-layer culture media filter, Fluorodyne® EX EDT, purposely designed to provide greater than 10 log removal of mycoplasma from complex soy-based culture media.

Since time and cost remain critical elements, users have become cognizant of throughput and value. “When biotechnology was in its infancy, users worried about whether sterilizing filters could sterilize, then they worried about validation. Today, those concerns have been allayed. Now, the big concern is filter capacity and their impact on manufacturing costs.”

Nowhere are cost and throughput concerns more evident than in virus filtration. This is understandable, Dr. Kumar explains, because “users focus on the things that cost the most, and virus filters are indeed the second most costly downstream step, right after capture chromatography.”

One could argue that the situation tilts even more strongly against virus filters than chromatography resins since the latter are generally reusable, whereas validation of virus filter re-use is problematic. One-time use only amplifies the negative perception. Virus filter cost, says Dr. Kumar, is justified by their expensive materials of construction and manufacturing technology compared, say, with sterilizing filters.

Pall recently introduced the Ultipor VF UDV20, a filter cartridge the company says will process more fluid at lower overall processing cost. VF UDV20 employs Pall’s laid-over pleating and narrow core cartridge technologies to pack 2 m2 of membrane into each 10 inch element. “What might have previously required two 20 inch cartridges can now be done in one 20 inch cartridge, at lower cost per liter.”

All things being equal, one would expect serum-free media to be cleaner and filter more easily than traditional media, but all things are not equal. “Serum-free media does tend to filter more easily,” says Mandar Dixit, head of product management for filtration technologies, North America at Sartorius Stedim Biotech. “But biomanufacturers have many additives that make the media they use filter differently from something off-the-shelf.”

Another confounding factor is high-temperature, short-time (HTST) treatment, which is increasingly common during media preparation. In some instances, HTST-treated media is more difficult to filter than nontreated media, Dixit says, “and since more and more biomanufacturers are incorporating heat treatment as part of platform manufacturing process, vendors have to conduct filterability trials at small and intermediate scales to confirm filter sizing done during process development.”

Sartorius Stedim is taking media filtration seriously, recently entering a collaboration with SAFC Biosciences to develop cost-effective, optimum filter trains for critical cell culture media for the firms’ common clients.

On its own, Sartorius Stedim Biotech  has introduced two additional multilayer PES membranes apart from its classical Sartopore 2 0.45/0.2 combination, to deal with evolving filtration needs. For example, the Sartopore 2 XLG, which was optimized for filtering soy-hydrolysate supplemented serum free media, features a 0.8 micron prefilter layer atop a 0.2 micron membrane; while the Sartopore XLI has 0.35 micron prefilter layer atop the same 0.2 micron final layer. Application-specific filter development represents the future of sterile filtration. Sartorius has also launched different double-layer PES prefilters, Sartoguard, to provide enhanced protection to its Sartopore 2 0.2 and 0.1 micron sterilizing-grade filters.  

Single-Use Still Gaining

Support for disposable bioprocessing remains one of bioprocess filtration’s ongoing trends. While significant resistance to single-use products remains for large, established processes, interest in development-stage manufacturing and upwards to approximately 1,000–2,000 liters is growing. That is the principal market for Spectrum Laboratories, which specializes in free-standing, disposable, hollow-fiber, tangential flow filtration (TFF) cartridges.

These products, says business development vp Dave Serway, form two families of products: one for process filtration, the other for perfusion cell culture. “Most similar products on the market require some sort of holder to keep the filter in place. Ours requires no stainless steel at all.”

Spectrum recently introduced a line of modified polyether sulfone hollow fibers, which are available at nominal molecular weight cutoffs between 10 and 500 kDa, with lumen diameters of 0.5 and 1 mm. Serway claims that on a volume-processed basis, these membranes are priced at 20–25% of conventional disposable cassette membranes.

Most top suppliers offer reusable TFF  cassettes. But according to Michael LaBreck, global product manager at Novasep, some end-users treat these as single-use products. Reasons vary, but the main justifications are the usual benefits of single-use vs. cost. LaBreck claims that his company has introduced the industry’s first purposely disposable TFF cassette that is priced significantly lower than reusable filters. “We have engineered in features for ease of use and process economics.”

The product, Sius™ Single Use TFF cassette, comes presanitized and ready to use out of the bag. “Traditionally, TFF cassettes are shipped in storage solutions consisting of glycerol and a bacteriocide. These fluids need to be washed out, and the filter sanitized, before use. Ours comes presanitized and preconditioned out of the bag, saving users costs associated with labor, water, and buffer.”

A typical cassette costs about $4,000 per square meter of filter area. Novasep claims a cost of $800 per square meter, with capacity and performance comparable to conventional TFF on a per-area basis. 

Novasep is targeting Sius Single Use TFF cassettes to clinical manufacturing and contract manufacturers. These users are typically running short campaigns of three to four runs and are particularly sensitive to costs. They can avoid costs for cleaning validations, reduce labor, reduce buffer usage and avoid the need for dedicated capital equipment by using a single-use strategy for TFF applications.

Vendors Meeting Diverse Needs

The uniqueness of filtration trains and specific products to individual processes becomes obvious from discussions with large bioprocess companies, particularly those who manufacture multiple products.

A large, diversified company like GlaxoSmithKline faces more challenges during the development of filtration unit operations than normal due to its unusual breadth of host organisms and molecule types. Still, Bruno Marques, Ph.D., an investigator at the company, notes that filter manufacturers have done a good job of addressing end-users’ needs. The evidence, he says, is in vendors’ responses to problems.

“Increasing titers made one thing crystal clear, that every unit operation along the purification process needs to achieve higher capacities and throughput.” As an example, Dr. Marques cites virus filtration. “Because of its relative high cost, improving virus filter capacity has been quite important to us. Top vendors have done a good job of addressing this issue, and of listening to our concerns.”

What could be improved? “Filter companies would do well to emulate chromatography resin companies and try to improve their scale-down models, eventually to achieve high-throughput methods for screening filter membranes.” Dr. Marques is referring to scalable, high-throughput methods that could be performed by robotic liquid-handling systems.

The problem with current filter scale-down models is that they are still in the 10–25 cm2 range for filter surface area. Each test consumes one gram or more of product, a price that Marques says is too high. “It would be great if at some point in the next few years biomanufacturers were able to plug scaled down versions of membranes into high-throughput systems.”

Sanofi-Pasteur, which manufactures influenza vaccine through an egg-based manufacturing process, relies primarily on ultrafiltration to remove lower molecular weight materials, and sterile filtration as a step important to antigen production and final product formulation. Success, says Joseph Frantz, Ph.D., director of pharmaceutical technology, depends on understanding and characterizing the material being filtered in light of available membrane types and materials that may be more or less compatible with the product.

Dr. Frantz calls filter product suppliers “excellent partners” for their willingness to conduct product compatibility studies. Other significant support areas include testing for integrity and Vmax, a parameter for estimating throughput and filter clogging. 

Sanofi-Pasteur recently employed presterilized, plug-and-play particulate filters for processes ranging from less than 100 L up to about 1,000 L and is working with suppliers to extend this methodology to sterile filtration systems. Disposables have been game-changers for the manufacture of small protein batches as well as for vaccines. Their success has been based not only on disposability and all its attendant benefits, but on connectivity and configurability. “When you purchase these products you’re not just buying a filter,” Dr. Frantz explains. “It’s critical to have the right sterile connectors and tubing, and with them the ability to plug the cartridge, aseptically, directly into the process.”