Recording date:

Duration:

Session:

Speaker:

Co-Authors:

Abstract:

Keywords

Tangential Flow Filtration, Microfluidics, Downstream Processing

Abstract

Sustainable vaccine manufacturing requires cost-effective, scalable, robust, and efficient processes. While process development efforts have largely focused on improving bioreactor yields, downstream yield is equally important since it can account for a significant contribution to the overall manufacturing costs. Key challenges in downstream processes are the optimisation of filtration and chromatography purification steps. For chromatography  evelopment, there are well-established microscale experimentation tools such as resin slurry plates and microscale columns to enable efficient screening studies. However, for tangential flow filtration (TFF), there is a lack of well-
understood and cost-effective small-scale screening devices that can identify optimal process parameters for achieving efficient filtration yields for industrial production. Microfluidic tangential flow filtration (µTFF) devices can be engineered to optimise filtration unit operations and reduce the cost of process development for large molecules downstream processing due to some of their key characteristics, such as low feed volume, laminar flow across
the system and the opportunity to integrate sensor technology. These characteristics will enable the study a variety of filtration products with different biochemical and physical properties, using µTFF devices as screening tools. However, there has not been enough efforts to bring these µTFF devices forward.

We have developed and characterised a novel parallel µTFF device for high throughput filtration process development. This device can be used to screen optimal conditions for the purification of viral vectors and vaccines. The prototype has been tested in concentration mode (5-10x with BSA feed concentration up to 10 g.L-1) at fluxes ranging between 29.6 LMH to 101.1 LMH (70 to 240 µl.min-1) at constant 15 psi transmembrane pressure (TMP). Results have shown no aggregation of protein during filtration with minimum retentate yields of 40%. Furthermore, robust operation was achieved (e.g. no internal leaks observed) which suggest that higher yields are achievable at low flow rates with constant TMP.

Integration with sensing technology allows the implementation of automated control strategies, removing operator-induced variability, therefore improving reproducibility and data quality. Our approach has the potential to deliver scale-relevant data, significantly reducing development times and process costs.