Final tangential flow filtration (TFF) formulation in combination with single-pass TFF (SPTFF) to reach 200 g/L of a monoclonal antibody using Delta 30 kDa TFF membranes
The final formulation of biopharmaceuticals, such as mAbs, is commonly performed using tangential flow filtration (TFF). This is often a two-step process with initial concentration using ultrafiltration (UF) followed by buffer exchange using diafiltration (DF) to the final formulation buffer. This process can take several hours with the mAb being recirculated in the TFF system at high pressures and shear forces. At the end of the process, the mAb is concentrated to its desired titer. However, during the final concentration, the risk of aggregation increases, hence the chemical composition of the formulation being critical to the stability of the mAb and its tendency to aggregate.
Despite the risk of aggregation, there are two main reasons for formulating the mAb at as high a concentration as possible:
- Some mAb treatments are administered to patients at high doses.
- More dilute formulations require more storage space and increase transportation costs, which has a greater environmental impact.
We have previously demonstrated a process to reach a mAb titer of above 200 g/L, see Production of a highly concentrated monoclonal antibody. In this study, we aimed to optimize the final formulation step using TFF and final concentration in a subsequent SPTFF step.
We used a T-series cassette with 30 kDa Delta regenerated cellulose membrane on an ÄKTA™ flux 6 TFF system. The final concentration was performed using a Cadence™ SPTFF modular kit with 30 kDa Delta regenerated cellulose membranes for as gen tle processing at lower shear. After the formulation, the final mAb product solution was 0.2 µm filtered using a Supor™ Prime filter.
Introduction
Final formulations of mAb solutions are performed through an initial UF concentration to reduce the volume followed by a buffer exchange using DF. This is usually carried out on flat sheet TFF membranes for mAbs. A final formulation process may take several hours and the mAb must withstand the shear forces and pressure being applied in the process.
One of the risks during a TFF final formulation is an increase in aggregate levels. After the DF, the mAb solution is often concentrated to its final titer. This may be a critical step since the volume will decrease with the risk of foaming the product when it is recirculated in the TFF system. The product will also become more viscous with an increase in shear stress, and the risk of aggregation will increase.
In this optimization, we ran a final formulation process for a mAb on Delta membranes of 30 kDa, 0.5 m2 using an ÄKTA™ flux 6 system as compared to running on Omega™ membranes of 30 kDa, 0.3 m2 previously. The change was part of the optimization as we knew from experience that Delta membranes are more suitable for mAbs at higher concentrations and we increased the membrane area to minimize the overall TFF process time for a gentler process.
We initially ran the UF to a titer of 50 g/L and then the DF buffer exchange was performed over 6 diafiltration volumes (DV) against a 50 mM acetate buffer at pH 5.0. The flux was monitored during the UF and DF. After the DF, no further concentration was performed through the TFF membranes. It should be noted that optimizing the formulation buffer to find a condition minimizing the level of aggregates during the TFF process has not been performed.
The final concentration was performed on a 4-in-series SPTFF unit, consisting of seven 30 kDa Delta membranes of 650 cm2 area in total, to minimize shear forces and risk of foaming. The final formulated product was filtered through a 0.2 µm Supor™ Prime filter membrane.
Our study aimed to optimize the TFF and SPTFF process steps to be as quick and gentle as possible while generating a highly concentrated final formulated mAb solution at 200 g/L with low aggregate levels. Another aim was to demonstrate that the SPTFF process could be run at stable pressures.
The process for this mAb is shown in green in Figure 1 below. However, for the optimization of TFF and SPTFF, we used the final formulated mAb (from the previous study) but the aggregate level was too high at 3.6%. Since most impurities were already removed, it was enough to reprocess on Capto™ S ImpAct ion exchange chromatography resin to reduce the aggregate level before the TFF and SPTFF steps.
The steps we performed in this optimization study are highlighted in blue in Figure 1 below.
Fig 1. An overview of the mAb purification process, with the TFF and SPTFF steps shown in blue, which were the focus of this article.
Materials and methods
Capto™ S ImpAct polishing purification
Before running the TFF- and SPTFF optimization according to Figure 1 above, we reprocessed the available final formulated mAb on Capto™ S ImpAct resin to reduce the level of aggregates. The polishing step was carried out as expected and the level of aggregates was reduced to 1.1%. We diluted the Capto™ S ImpAct eluate with the final formulation buffer to mimic the volume as if a full 50 L process had been run. It was important to have a representative start volume and start titer before the TFF step.
Fig 2. TMP and permeate flux during the UF and DF of the TFF step.
Final formulation through TFF
Before starting the TFF step, we rinsed the 30 kDa Delta membrane of 0.5 m2 and integrity tested at a pressure of 4 bar (58 psi, 0.4 MPa) with a passing result. We performed membrane cleaning in place (CIP) using 0.25 M NaOH followed by a water flush to remove the CIP agent. A normalized water permeability (NWP) test showed a value of 227 LMH/bar (at 20°C). We then primed the membrane with formulation buffer.
The diluted Capto™ S ImpAct eluate had a titer of about 7.2 g/L and the volume was 16.4 L. During the initial concentration, the volume was kept at about 6 L in the reservoir of the ÄKTA™ flux 6 system. Once all the product had been transferred, we reduced the volume to about 2.35 L. The UF part of the TFF step took about 33 min with a transmembrane pressure (TMP) of 1.5 bar (21.8 psi, 0.15 MPa) and a recirculation feed flux of 360 LMH. The following DF was performed with the same TMP and feed flux setting. An average permeate flux of 47.6 LMH was obtained reaching 6 DVs.
After the DF, we opened the retentate valve completely and the recirculation feed flux lowered to 120 LMH. We closed the permeate valve and the product was allowed to recirculate for 15 min to increase the recovery by gentle mixing and enhanced diffusion of mAb from the membrane surface back into the solution.
We collected the final formulated mAb retentate and the weight was 2156 g with a titer of 49.4 g/L and a total mAb content of 106.6 g. We performed two flushes of the ÄKTA™ flux 6 system with 300 mL of formulation buffer for each flush. The total flush was collected separately and weighed 592 g with a titer of 15.9 g/L and a total mAb content of 9.4 g. The total mAb amount was 116 g, which resulted in a step yield of 98.5% over the TFF step.
Both the retentate and the flush fraction were pooled with a weight of 2748 g, and a titer of 42.8 g/L. Before the final concentration on SPTFF, we filtered the product pool on a 0.2 µm filter.
The TFF membrane was cleaned in place after use and flushed with water. We performed a post-UF/DF NWP test, which gave a value of 229 LMH/bar (at 20°C) demonstrating full recovery of the membrane.
Figure 2 below shows the TMP and permeate flux during the UF and DF. Here we see that the permeate flux decreases as the product is concentrated during the UF. The drop in flux and TMP after about 33 min was due to lowering the recirculation flow for rearranging the buffer vessel and permeate collection vessel before the DF started. Once the DF started, we see that the permeate flux increases initially, which may be an effect of the viscosity change due to the buffer exchange. The TMP is seen to be stable during the UF and DF.
Final concentration through SPTFF
When running the SPTFF step previously (as stated in the article in the introduction), we used an 8-in-series unit. The process step took 7 h and the feed pressure was 3.8 bar (55 psi, 0.38 MPa). The step became uncontrolled, and we suspected that the 8-in-series was not fully utilized during this step. The starting concentration was also over 80 g/L and this was most likely too high to start with for an 8-in-series unit.
Optimization of this step therefore had to be performed and we assembled a 4-in-series SPTFF unit from individual Delta 30 kDa, 93 cm2 membranes, with a Cadence™ single pass modular kit instead. We controlled the retentate flow through a valve preventing unstable performance which most likely could lead to a too high concentration as the mAb concentration increases over the SPTFF unit. Another prerequisite was to start with a lower product concentration, and this was why we did not concentrate the mAb solution above 50 g/L in the previous TFF step.
Before starting the process, the assembled membranes in the 4-in-series SPTFF unit were cleaned in place with 0.25 M NaOH, flushed with water, and integrity tested at 4.0 bar (58.0 psi, 0.4 MPa) with a passed result and the following NWP test gave a value of 209 LMH/bar (at 25°C). The hold-up weight in the system was 45.5 g.
We carried out an initial flux investigation and the permeate and retentate lines were directed back into the feed container to investigate the optimal feed- and retentate pressures with the feed flow. The aim was to find the optimal point that results in a high concentration factor.
We carried out a few tests with a feed flux in the range of 85.7 to 15.7 LMH and feed pressures between 2.1 (30.5 psi, 0.21 MPa) and 3.6 bar (52.2 psi, 0.2 MPa) and retentate pressures between 0.70 (10.2 psi, 0.07 MPa) and 2.05 bar (29.7 psi, 0.21 MPa). The concentration factor observed varied between 2.2 up to 3.7. The optimal setting was finally found at a feed flux of about 16 LMH, feed pressure of 2.7 bar (38.4 psi, 0.27 psi) and retentate pressure of 2.05 bar (29.7 psi, 0.21 MPa). The volumetric concentration factor (VCF) then became about 5.3. These parameters were used for the SPTFF step.
Figure 3 below shows the flux/pressure excursion we performed to find the right settings.
Fig 3. Flux/pressure excursion at the start of the SPTFF step versus volumetric concentration factor (VCF).
The product pool had a weight of 2708 g and the run on the 4-in-series SPTFF using 30 kDa Delta membranes took 155 min at steady pressures. At the end, we performed three separate flushes with about 45 g of buffer for each flush. The titer in the concentrated mAb became 217.6 g/L after the SPTFF step. The product was finally filtered on Supor™ Prime membrane into a sterile bottle.
After the process, the SPTFF unit was cleaned in place and the following NWP test gave a value of 211 LMH/bar (at 25°C) showing full recovery of the membranes in the SPTFF unit.
Figure 4 below shows the accumulated retentate- and permeate weights throughout the SPTFF process, as well as the feed- and retentate pressure.
Fig 4. Accumulated retentate- and permeate weights throughout the SPTFF step as well as the feed- and retentate pressure.
Results and discussion
In this study we successfully optimized the TFF process for formulating a mAb solution on a 30 kDa Delta regenerated cellulose membrane of 0.5 m2 on an ÄKTA™ flux 6 system. Despite increasing the TMP to 1.5 bar (versus 0.7 bar [10.2 psi, 0.07 MPa] previously) for this process, we also increased the membrane area to 0.5 m2 (versus 0.3 m2 previously). This gave a faster process and hence the mab was exposed to high shear forces for as minimal time as possible. The gentler TFF process gave a lower increase in aggregates, from a start value of 1.1% to 1.3% (Fig 5).
We also showed that the SPTFF step could be optimized to give stable feed- and retentate pressures at the same feed flow rate during the entire process step. The processing time was reduced significantly to 2 h and 35 min (versus 7 h previously).
We also demonstrated a higher volumetric concentration factor of 5.3 when using the 4-in-series SPTFF module and this study confirmed that a 4-in-series SPTFF unit was sufficient for this step (as previously indicated and mentioned in the article referred to above in the introduction). The final titer in the product solution became 217.6 g/L and the aggregate level was at 2.6%. After the final Supor™ Prime filtration, the aggregate level was 2.4% and our specification limit was 3.0%.
The TFF and SPTFF steps gave high step yields of 98.5% and 98.0%, respectively.
Fig 5. mAb size variants through the TFF, SPTFF, and filtration step with Supor™ Prime membrane.
Conclusions
- The TFF process for the mAb solution on a Cytiva TFF membrane together with an ÄKTA™ flux 6 system was quick and led to a low increase in aggregate level over the TFF step.
- Final concentration above 200 g/L was achieved on an SPTFF unit with an optimized stable process on a 4-in-series SPTFF unit with Delta membranes generating a volumetric concentration factor of 5.3.
- The level of aggregates in the final product was at 2.4%, which met our final specification requirement.
Acknowledgements
This study was performed by Sandeep Kristiansson, Camilla Estmer Nilsson, Elon Svedlindh, and Olivia Lyngå. A special thanks to Gregor Kalinowski and Patricia Raleiras for assisting in the optimization of both the TFF and SPTFF step.
CY49686