Technical Forum: Disposable Chromatography

Disposables have been widely adopted for commercial-scale bioprocessing, but use of these technologies for downstream processing has lagged behind that for other applications. PharmTech spoke with industry experts about the challenges of implementing disposable chromatography systems.

Participating in the roundtable are Eric Grund, PhD, senior director of biopharma applications at GE Healthcare, Marc Bisschops, PhD, scientific director at Tarpon Biosystems, Tracy Thompson, CEO of Polybatics, Fred Mann, PhD, program manager, biopharm process solutions at Merck Millipore, and Stephen Tingley, vice-president, bioprocessing sales and marketing at Repligen.

IMAGE: INFLUX PRODUCTIONS, PHOTODISK, GETTY IMAGES

Barriers to implemetation

PharmTech: Chromatography has been one of the last components of the bioprocessing train to be adapted for single-use. What are the constraints of the chromatography process that have proved challenging to implement in single-use format?

Grund (GE Healthcare): The biggest constraint to single-use is probably a mental barrier based on a narrow view of the pros and cons. Chromatography media are often very tolerant to cleaning and withstand re-use, so it's tough to throw them away after single-use, especially if tests show they still perform well after many cycles. The benefits of speed, facility flexibility, facility output, and avoidance of cleaning are not yet fully appreciated.

Bisschops (Tarpon Biosystems): This statement is absolutely true for applications that involve capture of the product and/or some high resolution polishing steps. For flow-through applications (or negative chromatography), membrane adsorbers have already paved the way for disposable chromatography.

One of the most important reasons why chromatography has not been available in a disposable format is caused by the nature of the chromatography process itself: it is essentially a mass driven process, where the size of the column is governed by the amount of product that needs to be bound. For membrane processes and other flow-through applications, the most important system dimensions are determined by the volume that needs to be processed.

As a result, the successful introduction of disposable bioprocessing has largely been enabled by the process intensification that resulted from the increases in expression levels over the past decade. In essence, this has allowed us to produce the same amount of product with much less water and hence with a significantly reduced volume. All volume-driven unit operations have benefited from this, whereas the mass driven processes were not affected.

Everybody acknowledges that the costs of chromatography media currently are too high to justify a single-use application. These costs need to be depreciated over many cycles in order to make the economy work. This hampers the translation of batch-wise chromatographic processes into a single-use or disposable application, unless one uses a technology that would allow one to use the media over so many cycles in a single batch or in a campaign.

Thompson (Polybatics): Columns are very expensive systems, and the cost of buying these large chromatography systems is a cost that companies are reluctant to walk away from. Also, the cost of buying the resins themselves are fairly expensive, particularly Protein A. Protein A has been on patent until around March 2010, so there's been a monopoly on that particular ligand, which has maintained a very high price of the resin. I think those two factors have been a real impediment to going to a disposable chromatography system. And there hasn't been anyone out there who has come up with a format that is truly comparable to traditional packed-bed chromatography in terms of its ability to purify and capture the target.

In terms of implementation, packing of the columns is very fussy. You pump a slurry into the column, and have to let it settle. If it doesn't settle quite right, you can get voids in the column, and you have to pack again. There's a lot of art in packing the column to get it to perform right. One of the problems of implementing a disposable system is finding a medium that can either be pumped into fixed columns or finding a complete cartridge that is kind of plug-and-play. Until recently, there haven't been those kinds of plug-and-play systems.

Mann (Merck Millipore): Chromatography processes, while not fully single-use, have been operating in a hybrid way for some time with the implementation of single-use bags for buffers and product collection. Elimination of stainless-steel tanks and replacement with single-use bags is, together with the use of single-use bioreactors, the biggest contributor to cost savings when comparing single-use to traditional stainless-steel facilities. This is due to the elimination of clean-in-place (CIP) and steam-in-place (SIP) for tank/vessel cleaning. In contrast, the chromatography system is cleaned by process buffers including sodium hydroxide and does not need a separate CIP system.

Constraints of the chromatography process that make it difficult to implement as a disposable system include first, the greater complexity of the flow path in chromatography systems compared with other unit operations, for example the number of valves required to enable multiple buffer inlets, column flow reversal and bypass and fraction outlet. Coupled to this has been the greater number of different sensors deployed and the operating range and accuracy required of those sensors. Second, the cost of chromatography resins, especially the affinity resins such as protein A, has meant they tend to be used for multiple batches requiring cleaning and storage between times and so are not seen as single-use per se.

Tingley (Repligen): If we take a look at the process as a whole, and we look at the adoption curve of single-use technologies, you can essentially split the process into functional and nonfunctional technologies. It's the nonfunctional technologies that have taken the lead because they've been easier to implement and easier to get to an economical cost point than the functional technologies. Examples of nonfunctional technologies would be replacing stainless-steel pipework with plastic tubing, or replacing stainless-steel tanks with plastic bags. When you start looking at the process, for instance, a bioreactor or filtration technology such as ultrafiltration or microfiltration, these are examples of functional technologies, which have to be disposable. Making functional technology costs money, and functional technologies are often reused to defray some of the costs.

It just so happens that one of the most complex of the functional technologies is purification. That includes capture, using Protein A which we know is an extremely expensive chromatography resin, and hydrophobic interaction or ion exchange or multimode resins which are also reasonably expensive. And processes use a lot of them—that's multiple tens of liters multiplied by multiple thousands of dollars. With chromatography, it's a very expensive, very critical functional technology that is hard to get into a single-use format. So, there are two parts of the problem: can you make a disposable or single-use container for the chromatography, that is, a column, and then, can you make a single-use media or functional element to go into that. That's the problem that's made it so intractable.

When people want to move to single-use technologies, they may be reducing column sizes and cycling them harder. What users are doing is making the media work harder, so it's less painful to throw it away. What you're seeing today is companies offering the easy part, the containment part, of the disposable chromatography, the column shells, and packing them. The difficult part of the technology is finding new ways to stretch the economics of running longer, running smaller batches, cycling the columns more often, and things like that.

Choosing a disposable platform

PharmTech: A few disposable chromatography platforms are currently available, including packed-bed, simulated moving bed (SMB), and membrane chromatography. What are the factors that would influence the choice of platform for a process?

Grund (GE Healthcare): Packed beds are used in steps following feed clarification, when binding capacity and resolving power are prioritized. Conventionally, the first step in downstream purification is product capture, in bind/elute mode, and a packed bed is needed to achieve the objectives of the unit operation.

The question of whether or not to use SMB is different. Frequently, a small number of cycles is used to handle large volumes of feed. SMB takes this further by providing a continuous processing approach with several small columns cycled in sequence. Generally, SMB offers higher loading capacity, greater exploitation of resin life, and more efficient use of buffers. So, SMB can be a door-opener to using disposable chromatography columns because small columns are used for multiple cycles to handle material from the bioreactor. This helps address the cost equation because the resin is used for many cycles before disposal. Accurate control and synchronization of the different phases in the chromatography cycle is critical. Single-use components are attractive in SMB since they assure reproducible performance and avoid multiple column-packing in the production work-flow. The downside in SMB is system complexity. Multiple columns require many valves and sophisticated control to assure accurate column switching without cross contamination.

Chromatography is also used in product flow-through mode to remove impurities. The further downstream you are in your process, the fewer the impurities. When there are only small amounts remaining, membrane chromatography is attractive for impurity scavenging. The low binding capacity and low resolving power is not an issue and the high flow rates that can be used can be fully exploited. Single-use components are often preferred for scavenging, especially because cleaning may be challenging for several reasons.

For producing tons of product, the column volumes are large, several hundred or even thousands of liters, and at this size, single-use designs are not viable. In general, the smaller the scale, the more attractive single-use chromatography is.

Bisschops (Tarpon): First of all, disposable technologies will generally result in more flexibility in manufacturing and in a shorter change-over time. These features are particularly important for multiproduct facilities such as clinical manufacturing facilities and contract manufacturing organizations. For these types of operations, the advantage of disposable bioprocessing technologies is more obvious than for single-product facilities.

Prepacked chromatography columns fit very well in streamlining the workflow in a facility by taking away the packing operations. The costs for prepacked columns were, until recently, cost prohibitive to consider them as a single-use or disposable product for other applications than clinical manufacturing. The scale limitations of prepacked columns also restricts the application of this technology to clinical-scale manufacturing.

SMB enables manufacturing of large amounts of product with reasonably small columns, which are cycled many times during a batch. As a result, this technology can make prepacked disposable chromatography a viable alternative, especially when you pair disposable columns with a fully disposable, simplified valving system. Disposable valving is the missing link in providing economically viable, fully disposable downstream processing for bind/elute applications. Another feature of continuous chromatography is that it allows the entire cascade of downstream processing unit operations to be operated as a fully continuous train. This continuity eliminates or significantly reduces interstage product hold steps and allows multiple unit operations to be operated simultaneously. Thus, the time in facility can be shortened by a factor of two, which in many cases translates into a significant increase in facility throughput.

Mann (Merck Millipore): Probably, the specific application is the first criterion, in so much as to how much freedom there is to pick and chose a chromatography platform. For instance, if the product is a monoclonal antibody, then likely there is a template already in place for purification, typically protein A affinity chromatography, followed by cation exchange bind-elute and then anion exchange flow through. Both the protein A affinity and cation exchange are almost certainly going to be conventional packed-bed columns. Although traditionally the anion exchange was also a packed-bed column, anion exchange flow through membrane adsorbers are being deployed because of convenience (i.e., no column packing) and buffer savings (i.e., no cleaning/reuse).

Membrane adsorbers are also finding application elsewhere when used in flow through mode for capture of impurities. Generally, they are less competitive with conventional packed columns for bind-elute applications because of lower capacity compared with resins.

SMB is relatively new to biotech. While frequently used for small molecule purification, it has not found adoption in protein separations primarily due to the greater complexity of the flowpath and the difficulty with engineering it in a sanitary manner. The new single-use systems coming onto the market may address that aspect, but the added complexity of operation compared to conventional batch chromatography will likely continue to be a hurdle to adoption. One attraction of SMB or similar multicolumn approaches is that, compared with batch, it uses smaller columns that make it more amenable to prepacked columns and coupled to the fact that the resin is cycled more times per batch. This has benefits especially for clinical-scale batches where, conventionally, the resin may be thrown away after only a few batches and so is nowhere near its end of life point. Multicolumn approaches enable better resin utilization, getting closer to the lifetime of the resin and thus saving cost.

The second criterion is probably scale, which is linked to cost. Although single-use implementation shows clear cost benefits at the smaller pilot/clinical scale manufacturing, at large commercial scale, stainless-steel installation can be more cost-effective. In addition, larger scale will require larger columns than currently available in a prepacked, disposable format.

Tingley (Repligen): In stepping back a little bit, you can ask—why do people want to adopt single-use technologies? I don't think the answer has changed as we've changed the technologies—it's speed—getting through the process quicker, being able to develop multiproduct facilities, being able to put more molecules through a facility in a short period of time. This is the reason why the disposable trend has developed and has been so successful over the past 15 or 20 years.

When you look at chromatography and ask what do people want to do with a disposable system, there are two answers. At one end of the continuum are users who really want to use chromatography columns the way they've always used chromatography columns, they just don't want to pack them any more. At the other end of the continuum are companies that have built truly single-use platforms, and don't have any capability to manage hardware. These companies are buying disposable columns and they're throwing away the columns and the media, because although on a per-process basis it may seem expensive, in terms of the overall operation, they get the economy of scale. Users operate on a continuum, and you see systems ranging from pure disposables to hybrid facilities. So, when you ask what factors influence the choice of platform, the answer for me is that the traditional process and the traditional technologies are the number one drivers. How do people get the convenience of disposables doing what they do today?

In addition to packed columns, there are also alternate technologies such as membranes that are good for some of the flow through applications. I think it's still early days, but some of these technologies are working well and tend to be in smaller processes.

Barriers to adoption

PharmTech: What barriers do you see to more widespread adoption of single-use chromatography?

Grund (GE Healthcare): Weighing up the pros and cons will give a different answer from case to case and it really depends on the application in question. Not all applications are best suited to disposable chromatography columns, speed is not the only goal. Operational efficiency can be addressed in other ways and many hybrid solutions are possible. Another factor is obviously that large, hard-piped facilities in existence around the world—dedicated to a small number of products—can operate very economically and there is little motivation to refurbish them. There are also many smaller facilities based on conventional stainless-steel approaches and these will probably only be replaced with single-use components when the pressure on flexibility and facility throughput is high.

The strongest push towards single-use is in multiproduct facilities that switch products frequently, especially at small scale, for example, in process development, production for clinical trials, or contract manufacturing. Here the barrier is more related to a conservative attitude with general reservations about using disposables. Manufacturers are concerned with issues such as risks from leachables, poor documentation, and increased risk of operator error.

Bisschops (Tarpon): One of the most important barriers to introducing disposable chromatography is most likely the sunk capital in legacy facilities.

We do see, however, a growing trend in even existing legacy facilities moving to disposables in process steps where the facility design itself becomes a limitation either because of increasing titers, holding tank capacity, or water-for-injection capability. This is where disposable continuous processing can have a huge impact.

Thompson (Polybatics): I think one of the challenges that equipment suppliers haven't really addressed is the cost issue. Single-use manufacturers haven't really addressed the cost aspect—they've just shifted them from one-time upfront to ongoing operational expenses. I think you have to get to a 30% savings or more before manufacturers will take the investment they have in existing processes and systems and shift them. Otherwise, there just doesn't seem to be the economic incentive to shift to disposables.

Will there ever be a completely disposable system? There will still be some components of any system, whether it's membrane or resin-based, that will be reusable. I do believe there will be systems on the horizon as technologies evolve that are truly disposable. Whether that means systems that are single-use, or say, 10 uses for a campaign... My suspicion is that it will be more along the lines of using a unit for a campaign, then once the campaign is done you get rid of it. So you still get the benefits of disposablility but leverage some of those costs over 10 or 20 cycles.

Mann (Merck Millipore): One barrier is the availability of systems, including systems that have gradient capability. Greater capability and system choice will likely drive adoption.

A second barrier is the availability of true single-use devices or prepacked columns. While single-use membrane adsorbers are being adopted, especially for flow through applications, the relatively lower capacity compared with resins limits application for bind-elute applications. Consequently, higher capacity membrane or similar device format could facilitate adoption. Alternatively, or in combination, lower cost, prepacked columns would make single-use operation more attractive.

Tingley (Repligen): For me, the way to get adoption of these chromatography products is to make them easy to adapt to what people are doing now. If they can get a prepacked column that's prepared in exactly the same way as their glass column, that's used exactly the same way as their glass column, and gives exactly the same results as their glass column, that will be the first step in making chromatography disposable. With that, I think, will come pressure to look for alternatives to make chromatography truly single-use.

From a vendor's point of view, for years we've been having great conversations with the biopharma industry about introducing new technologies and making changes. But, the actual adoption rate of game-changing technology is poor. Just because of the way we do things in this industry, we're more likely to be evolutional than revolutional. As long as people can think about how they can use the product as a disposable or semidisposable step and it makes sense to them, then they can easily see it fit within the confines and contraints of their own company's regulatory philosophy and guidance, and it's an easy step to make. This is what will take us down the path to truly game-changing technology in the years to come.