Challenges in surfactant removal from biopharmaceutical formulations using tangential flow filtration (TFF) and spin columns

Abstract

Surfactants such as polysorbate 20 (PS20), polysorbate 80 (PS80), and poloxamer 188 (P188) are integral to the stabilization of protein therapeutics, yet their removal is often required during formulation development and analytical characterization. This work investigates the efficiency of surfactant removal by commercial spin columns and tangential flow filtration (TFF), using liquid chromatography with charged aerosol detection (LC-CAD) to evaluate subspecies specific behavior. Spin column experiments demonstrated effective PS20 clearance from monoclonal antibody formulations, while PS80 removal was limited, particularly for PS80 polyester species. TFF enabled selective removal of unesterified and monoester species for both polysorbates, with PS20 showing the highest overall clearance, especially when processed below its critical micelle concentration (CMC). PS80 removal was ineffective again due to retention of polyester species, while P188 exhibited slower and incomplete removal kinetics. These findings underscore the need for improved surfactant removal strategies and the importance of subspecies profile analysis in surfactant quantification.

Introduction

Non-ionic surfactants, in particular polysorbate 20 (PS20), polysorbate 80 (PS80), and poloxamer 188 (P188), are key stabilizers for biologics against adsorption to interfaces and surface-induced unfolding and aggregation.1,2 However, once surfactants are added to a formulation, it is highly challenging to remove them again. During biopharmaceutical development, multiple scenarios would benefit from a quantitative surfactant removal approach, for example i) early formulation screening studies with surfactant-containing starting material, ii) biosimilarity studies comparing the biosimilar and originator in standard formulation conditions, and iii) the validation and benchmarking of novel in vitro and in silico analytical methods by analysis of commercial drug products, particularly for early-stage screening.3 Additionally, several analytical methods are susceptible to matrix effects or system deterioration when applied to surfactant-containing samples.4,5 Sample volumes to be cleared of surfactants may range from a few microliters for analytical purposes up to several liters of drug substance for formulation studies, requiring tailored approaches and methodologies. Due to the inherently amphiphilic nature of surfactant molecules, their chemical heterogeneity and frequent usage above the critical micelle concentration (CMC),2 quantitative surfactant removal represents a significant challenge.6 The behavior of commonly used surfactants and their respective subspecies (polyesters, monoesters and unesterified species for polysorbates; main and early eluting species for poloxamers) in various removal methods is not yet fully understood. Previous studies of surfactant removal by dialysis7 and centrifugal concentration8 suggest low overall removal efficiencies or even surfactant retention and concentration. For other removal approaches, our understanding is even less comprehensive. Lei et al. attempted PS20 removal by tangential flow filtration (TFF) and showed partial removal.9 Various surfactant removal spin columns offer a convenient commercial solution for some surfactants, but their performance has not been fully investigated.10

In this work we present three surfactant removal case studies performed during independent biopharmaceutical development projects, with the aim of highlighting the difficulties of achieving quantitative surfactant removal using commonly used methods. We investigated the use of two commercially available surfactant removal spin columns as part of efforts to reformulate various monoclonal antibody (mAb) drug products (Table S1). These centrifugal columns are frequently used in early development settings to process small volumes (typically up to 10 mL). In addition, we also tested PS20, PS80, and P188 removal by TFF as a technology potentially capable of processing larger volumes. Subsequent surfactant quantification was performed by using a liquid chromatography with charged aerosol detection (LCsingle bondCAD) method we previously developed.11 Notably, the chosen method achieves chromatographic separation of surfactant subspecies, which provides additional insight into the removal behavior of multiple commonly used surfactants beyond total quantification.

Continue reading here

Kristian Le Vay, Benjamin Steinborn, Constanze Helbig, Andrea Arsiccio, Maksymilian M. Zegota, Christa von der Schulenburg, Angelika Reichel, Tim Menzen, Andrea Hawe, Challenges in surfactant removal from biopharmaceutical formulations using tangential flow filtration (TFF) and spin columns, Journal of Pharmaceutical Sciences, Volume 115, Issue 2, 2026, 104141, ISSN 0022-3549, https://doi.org/10.1016/j.xphs.2025.104141.

Read more on Biopharmaceutical Projects here:

• Machine learning driven acceleration of biopharmaceutical formulation development using Excipient Prediction Software (ExPreSo)
• A smart solution for Biopharmaceutical Development – by Hayashibara
• On the role of excipients in biopharmaceuticals manufacture: Modelling-guided formulation identifies the protective effect of arginine hydrochloride excipient on spray-dried Olipudase alfa recombinant protein