What is Continuous Chromatography?
Continuous Chromatography is a new process for manufacturing biomolecules, primarily antibody drugs. The biopharma industry is facing increasing demand for large quantities of recombinant proteins due to the growth in the protein drug market as well as advances in proteomics creating the need for faster and more rugged manufacturing processes. Historically, large scale manufacturing for pharmaceutical has relied on a single column chromatography recovery processes, but is now moving increasingly to Continuous Chromatography methods.
Continuous Chromatography is a semi-continuous process that drives a primary column to complete loading equilibrium while capturing overflow on a secondary column next in line. Subsequently, when loading is complete, the primary column is removed. The former secondary column becomes the new primary column with a new secondary column installed in series to catch overflow. The removed primary column is washed to remove impurities and eluted to recover the drug protein product. After that, the column is cleaned and conditioned to make the column ready for reuse. Typically a column resin packing is reused about 200 times and then the column is repacked with new resin.
In Continuous Chromatography, the loading of the column is based on column selectivity for the drug protein under the loading buffer conditions that are applied. Loading is not based the kinetic rate of interaction, flow rate linear velocity, packing uniformity or packing diameter. The column is loaded to equilibrium. The column is loaded to equilibrium, regardless of flow rate and break through profile.
What does “loaded the column to equilibrium” mean and why is that important?
Even if the breakthrough flow profile is not sharp, the overflow material is not lost. Overflow material is captured on a second column that is held in series with the first column. This means that the column is completely loaded because the chemistry of interaction is brought to completion and equilibrium. The column equilibrium is the result of resin selectivity and the buffer conditions that are used.
Other reasons on what this means and why it is important will be explained later in this Q&A in the discussion on Dual Flow Chromatography.
How is this different than the single column process currently in use?
In a single column process, the manufacturing column is loaded as rapidly as possible using a rapid flow rate while still maintaining a sharp sample breakthrough profile. Loading is complete when the drug just starts to break through the column. If the column breakthrough profile is sharp and the column is loaded to the resin at the outlet end of the column, this column is also said to be completely loaded and at loading equilibrium. After loading, the column is washed, eluted to recover material and then reconditioned to start the process again.
For drugs that have a rapid kinetic uptake (and a sharp, flat breakthrough curve) high flow rates may be used. But drugs with slow loading kinetics, the capture profile may not be sharp and product could be lost in the overflow. In these cases, flow rate must be lowered to sharpen the breakthrough curve, albeit at a penalty of increased manufacturing process time.
So loading to completion and to equilibrium is possible with a single column system. However, in these cases, the flow control is important to maintain a sharp profile. The velocity required will vary with the kinetics of capture of the individual proteins of interest.
Why is the market shifting?
Since the capture kinetics of every protein is different, optimizing a method to manufacture a drug with a single column takes rigorous development of the flow rate and close monitoring of the column during operation. Failure to do this results in valuable product being lost.
It is now possible to err on the side of having too fast flow rates without losing product. In addition, capture is now a continuous process – there’s no interruption. Both of these vastly increase the speed of manufacturing. Productivity has essentially more than doubled.
But isn’t it more expensive to use a continuous process?
Capital hardware costs are higher. Yes, more switching and control hardware and software are needed. More manufacturing columns are needed. But column hardware costs of the individual columns can be decreased because smaller columns can be used.
The overall operating costs, including the resins costs, are identical – it is not more expensive. No additional resins and buffers are needed, so there are no additional consumable costs per unit of delivered product.
The labor costs are lower because productivity of producing the final product is higher.
Are there any other benefits to Continuous Chromatography?
Yes, it turns out there are several somewhat unexpected benefits. At the conference, PhyNexus introduced a new chromatographic concept: Dual Flow Chromatography. The paper presented was “Dual Flow Chromatography for Parallel, Automated, High Throughput Process Development”
Dual Flow Chromatograph (DFC) is a process where separations are performed with back-and-forth flow of the samples and the mobile phase through pipette tip based columns. Typical column bed sizes are very small, ranging from 5 to 160 µL. Although different than conventional unidirectional flow-through chromatography, chromatographic principles still control the dual flow process. Capture of sample is brought to equilibrium with back and forth flow through the column; the flow is the sample as pipetting the solution several times. The back and forth cycle is completed until capture is complete – usually in 4 to 6 cycles. Separations are performed 96 at-a-time using standard laboratory liquid handing robots.
How does Dual Flow Chromatography benefit Continuous Chromatography?
Although the flow dynamics are different, both Continuous Chromatography and Dual Flow Chromatography operate on the principle of bringing column resin and sample molecule interaction to equilibrium and completion. In both methods, sample molecule capture is complete based on resin and buffer conditions, regardless of sample molecule kinetics.
But how does that make it beneficial?
First, because it makes it possible for Dual Flow Chromatography to be used to predict the quantity and the quality of a product that can recovered from a column regardless of the column size. Small columns can be used to predict the performance of preparative and manufacturing columns. In the paper presented, a plot of amount of protein captured vs. the column size was linear with a slope of one. Scaling can be accomplished regardless of column size, provided the columns are completely loaded. This is will be discussed further in the Q&A.
Second, because the separation conditions developed in parallel under automated conditions can be used to predict the performance of a continuous chromatography column operated under the same buffer conditions. Multi-variable conditions can be tested quickly and compared directly. Literally thousands of methods with different resins, capture, washing and elution conditions can be tested.
Existing method development & optimization techniques such as plates and FPLC are not keeping pace with advances in manufacturing scale production. Lab scale method development is slow, and focuses primarily on protein mass capture (yield) and, to a lesser extent, protein purity.
Lab processes are not designed for efficient multi-variate, multi-parallel experimentation for yield, purity and protein activity. Lab processes may not include protein activity or agglomeration measurements – an absolutely critical measure of ultimate effectiveness at large scale.
Why is multi-variate screening important?
Proteins are complex molecules. Agglomeration of proteins can occur at any step in the purification process. Proteins may denature. Recovery must be high. The proteins must remain denatured and active. Multi-variate screening of conditions including buffer type, pH, concentration, salt additives, surfactants, etc. is the only possible way all these goals can be achieved.
Dual Flow Chromatography can be used to develop and evaluate literally thousands of possible manufacturing procedures within a matter of days – all done in miniature columns and processed in parallel on automated robotics.
In one DFC experiment described in the paper presented, 288 purifications conditions were screened for 4 clones – 72 different procedures for each clone. The work was completed within a day. Yield was measured by ELIZA and protein activity measured by SPR. The result showed that it is possible to have high yield and very low activity. But success is difficult and only if the correct conditions are used. In fact according to these experiments, it is much more likely that only high yield is achieved or high activity is achieved, but not simultaneously.
Other references where DFC multi-variant method development was used for scale-up: R. Hopkins, D. Esposito, and W. Gillette “Widening the bottleneck: increasing success in protein expression and purification” J Struct Biol. 2010 October ; 172(1): 14–20., M. D.Wenger, P. DePhillips, C. E. Price and D. G. Bracewell “An automated microscale chromatographic purification of virus-like particles as a strategy for process development” Biotechnol. Appl. Biochem. (2007) 47, 131–139 and Michael Rauscher, John Welsh, Jennifer Pollard Process Development and Engineering, Bioprocess Development, Merck & Co., Inc., Kenilworth, NJ, USA “Application of High Throughput Resin Tips for Chromatography Development” poster PREP 2015, The Loews Hotel, Philadelphia, July 26-28, 2015.