Prediction of viral filtration performance of monoclonal antibodies based on biophysical properties of feed

• Published in: Biotechnology progress Vol. 31; no. 3; pp. 765 - 774
• Main Authors: Rayfield, William J., Roush, David J., Chmielowski, Rebecca A., Tugcu, Nihal, Barakat, Shehab, Cheung, Jason K.
• Format: Journal Article
• Language: English
• Published: United States Blackwell Publishing Ltd 01.05.2015; Wiley Subscription Services, Inc
• Subjects: Antibodies, Monoclonal - chemistry; Antibodies, Monoclonal - isolation & purification; Biophysics; Buffers; Chromatography; Drug Contamination; dynamic light scattering; Filtration; Hydrogen-Ion Concentration; Ions; Molecular Weight; monoclonal antibody purification; Spectrum analysis; viral clearance; virus filter performance; virus removal filters; Viruses - isolation & purification; dynamic light scattering; monoclonal antibody purification; viral clearance; virus filter performance; virus removal filters
• ISSN: 8756-7938; 1520-6033
• DOI: 10.1002/btpr.2094

Controlling viral contamination is an important issue in the process development of monoclonal antibodies (MAbs) produced from mammalian cell lines. Virus filtration (VF) has been demonstrated to be a robust and effective clearance step which can provide ≥4 logs of reduction via size exclusion. The minimization of VF area by increasing flux and filter loading is critical to achieving cost targets as VFs are single use and often represent up to 10% of total purification costs. The research presented in this publication describes a development strategy focused on biophysical attributes of product streams that are directly applicable to VF process performance. This article summarizes a case study where biophysical tools (high‐pressure size exclusion chromatography, dynamic light scattering, and absolute size exclusion chromatography) were applied to a specific MAb program to illustrate how changes in feed composition (pH, sodium chloride concentration, and buffer salt type) can change biophysical properties which correlate with VF performance. The approach was subsequently refined and expanded over the course of development of three MAbs where performance metrics (i.e., loading and flux) were evaluated for two specific virus filters (Viresolve Pro and Planova 20N) during both unspiked control runs and virus clearance experiments. The analyses of feed attributes can be applied to a decision tree to guide the recommendation of a VF filter and operating conditions for use in future MAb program development. The understanding of the biophysical properties of the feed can be correlated to virus filter performance to significantly reduce the mass of product, time, and costs associated with virus filter step development. © 2015 American Institute of Chemical Engineers Biotechnol. Prog., 31:765–774, 2015