High throughput screening of ultrafiltration and diafiltration processing of monoclonal antibodies via the ambr® crossflow system

Abstract

As the biopharmaceutical industry moves toward high concentration of monoclonal antibody drug substance, additional development is required early on when material is still limited. A key constraint is the availability of predictive high‐throughput low‐volume filtration screening systems for bioprocess development. This particularly impacts final stages such as ultrafiltration/diafiltration steps where traditional scale‐down systems need hundreds of milliliters of material per run. Recently, the ambr® crossflow system has been commercialized by Sartorius Stedim Biotech (SSB) to meet this need. It enables parallel high throughput experimentation by only using a fraction of typical material requirements. Critical parameters for predictive filtration systems include loading, mean transmembrane pressure (ΔP¯ TMP), and crossflow rate (QF). While axial pressure drop (ΔPaxial) across the cartridge is a function of these parameters, it plays a key role and similar values should result across scales. The ambr® crossflow system is first presented describing typical screening experiments. Its performance is then compared to a traditional pilot‐scale tangential flow filtration (TFF) at defined conditions. The original ambr® crossflow (CF) cartridge underperformed resulting in ~20x lower ΔPaxial than the pilot‐scale TFF flat‐sheet cassette. With an objective to improve the scalability of the system, efforts were made to understand this scale difference. The ambr® CF cartridge was successfully modified by restricting the flow of the feed channel, and thus increasing its ΔPaxial. Additional studies across a range of loading (100–823 gm−2); ΔP¯ TMP (12–18 psi); and QF (4–8 L/min/m2) were conducted in both scales. Comparable flux and aggregate levels were achieved.

Streamlining process characterization efforts using the high throughput ambr® crossflow system for ultrafiltration and diafiltration processing of monoclonal antibodies

“…This paper will initially compare the UF/DF performance of the ambr ® crossflow, a pilot-and a manufacturing-scale TFF system at a single condition. It will then extend the previous work, 19 and compare the performance of the ambr ® crossflow with the pilot-scale TFF system across a 3-factor DoE design space used in a formal PC study.…”

“…Each channel is fitted with an inline pH and conductivity sensors in the permeate line, and pressure sensors in all lines which are all automatically calibrated prior to each experimental run. All experiments in this article use a modified ambr ® crossflow cartridge, 19 whereby the axial pressure drop was increased by a physical modification of the cartridge interior design. All ambr ® crossflow experiments were conducted using the same initial volume (i.e., $55 ml totaling to a loading of 600 gm −2 ) and final retentate volume (i.e., $8 ml) were used.…”

“…A traditional pilot-scale TFF system fitted with a feed/retentate tank, crossflow pump (QuattroFlow™ 150S, Triangle Process Equipment, Wilson, NC), diafiltration pump (Masterflex ® L/S™, Cole-Parmer Instrument Company, Vernon Hills, IL), 0.0088 m 2 membrane cassette with holder, and automatic retentate pinch valve was used in these experiments (Figure 1b), which is further described elsewhere. 19,20 Single-use pressure sensors (PREPS-N-025, PendoTECH, Princeton, NJ) were located in the feed, retentate and permeate line.…”

Strategies for high‐concentration drug substance manufacturing to facilitate subcutaneous administration: A review

“…This can be evaluated by testing samples before and after high shear unit operations carried out under worst‐case conditions within the parameter operating space (e.g., highest loading, longest run times, conditions leading to largest number of pump passes). The selection of appropriate process parameters for UF/DF may also be supported through high‐throughput screening tools, such as the ambr® crossflow (Sartorius Stedim Biotech; Fernandez‐Cerezo, Wismer, Han, & Pollard, 2019) or a multistation PendoTECH screening system (Bhangale, Chen, Khetan, & Furey, 2018). Since high‐throughput screening tools often have lower material requirements than traditional lab‐scale systems, they can provide a particular benefit for high‐concentration development work where large volumes of development material are often limited.…”

“…Although the material requirements for the screening systems are lower than traditional lab‐scale systems, they can still be substantial (e.g., 0.5–1 L), particularly during early development where material is often very limited. A key challenge is the availability of scale‐down membrane formats that are representative of the flux and axial pressure drop observed at the larger scales (Fernandez‐Cerezo et al, 2019). While the screening studies can provide information about some scalable membrane characterization properties (e.g., membrane rejection or permeability), the hydrodynamic properties (e.g., shear rate, pressure drop, and mass transfer) are influenced by edge effects and boundary layer formation which are difficult challenges to overcome in representative scale‐down devices.…”