Development and validation of a selective marker-based quantification of polysorbate 20 in biopharmaceutical formulations using UPLC QDa detection

Highlights

• Development of a UPLC QDa method to quantify polysorbate 20 components.

• More than 650 different polysorbate 20 components were identified.

• Selected polysorbate 20 components were defined as markers.

• Clever sample preparation and calibration procedure.

• Successful validation of the method at different polysorbate 20 levels.

Abstract

Polysorbates are widely used as non-ionic surfactant in biopharmaceutical formulations. Recently, the degradation of polysorbate moved into the focus of attention, because in several published studies it was described, that stability issues in polysorbate containing formulations were observed leading to the formation and appearance of sub-visible and visible particles. For this reason, monitoring of polysorbate and its degradation products is of importance throughout the development of parenterals. The aim of the study was to develop a method for the selective marker-based quantification of adequate polysorbate 20 components of interest without the need to apply derivatization or complex detection techniques. A single quadrupole mass (QDa) detector was used coupled to an ultra-high performance liquid chromatography (UPLC) system.

Method development was based on a reversed phase-high performance liquid chromatography assay coupled to a charged aerosol detector (RP-HPLC CAD). Instead of a charged aerosol detector (CAD) a QDa detector was used in order to significantly improve the selectivity. The focus of this study is the development of the QDa based method for the analysis of polysorbate 20. Modifications of the mobile phase and the type of chromatography column allowed the separation of several components of polysorbate 20 from polar non-esterified to apolar higher order species. In addition, a multitude of components could be quantified by their individual m/z values. The peak assignment identified 676 compounds which originated from polysorbate 20. Some of these were selected and defined as marker components. It was shown that the developed method is capable to determine polysorbate 20 in different biopharmaceutical formulations. The proposed assay is based on a smart sample preparation as well as a unique calibration procedure that make the determination of several selected components achievable. Furthermore, it was successfully demonstrated that the analytical procedure is valid to reliably quantify several polysorbate 20 components at its 100% level (corresponds to 0.4 mg/mL intact polysorbate 20) and even at lower concentrations that occur e.g. in case of polysorbate 20 degradation. In conclusion, the method is beneficial to determine selected polysorbate 20 species during formulation development of biopharmaceuticals as well as during stability testing and trouble shooting.

Introduction

Non-ionic surfactants like polysorbates are often used in biopharmaceutical formulations to stabilize the protein against several environmental stress conditions. Polysorbate 20 and polysorbate 80 are widely used in parenteral formulation development, because of their biocompatibility and their low toxicity [1], although the stabilization mechanism(s) for biologics is unclear [2], [3].

According to the Ph. Eur. polysorbates are a heterogeneous mixture of partial esters of fatty acids with sorbitan and isosorbide which are further ethoxylated with polymerized ethylene oxide [4], [5]. The chemical synthesis starts with the dehydration of sorbitol to form a mixture of sorbitan and isosorbide followed by the addition of fatty acids and their subsequent esterification. Finally, the mixture is ethoxylated with OE (oxyethylene) units to form sorbitan and isosorbide alkyl esters [6], [7]. For polysorbate 20 at least 9 different fatty acids are listed in the Ph. Eur. [8].

Recently, the degradation of polysorbate has emerged as a major challenge in the biopharmaceutical community. Two major degradation mechanisms are known: hydrolysis and oxidation [9]. While oxidation is likely in biopharmaceutical formulations and might be reduced by addition of radical scavengers and/or changing polysorbate quality [10], [11], hydrolysis should not be expected under pharmaceutical relevant solution conditions [12]. However, the presence of traces of host cell proteins like esterases as residues from the drug substance manufacturing process can lead to hydrolytic degradation of polysorbates [13], [14], [15]. Currently, the enzyme(s) are not really identified; a few, specific enzymes were reported in the literature, among these are discussed PLBL₂, LPL, phospholipase A2, LPLA₂, carboxylester hydrolases, that have been shown to degrade polysorbate [13], [14], [15], [16], [17], [18], [19], [20]. Dependent on which degradation mechanism prevails, the degradation products are various. Non-esterified ethoxylated sorbitans and isosorbides and corresponding free fatty acids (FFA) mostly occur through hydrolysis. Zhang et al. (2017) have observed similar degradations due to oxidation of all-laurate polysorbate 20, namely the formation of free lauric acid and polyoxyethylene (POE) [21].

Depending on their water solubility and the amount of intact polysorbate present, the typically poor water soluble FFA can finally phase-separate leading to the appearance of sub-visible and visible particles [22], [23], [24], [25]. The appearance of FFA might also impair protein stability. Other degradation products particularly formed during oxidation are aldehydes and ketones which might also influence the stability of the formulation, such as the formation of aldehyde adducts [26]. For this reason, monitoring of polysorbate and its degradation products is of importance throughout the development and, stabilization of biologics. Consequently, the need of a reliable quantification method for polysorbate is enormous and very challenging due to the complexity of the mixture. A number of methods were developed in the last decades. Most of them are capable to screen several polysorbate components and assess the polysorbate composition qualitatively or determine a defined component of polysorbate as reference marker of quantification [27]. One of the first and well-known method is the quantification of polysorbate by fluorescence micelle assay (FMA). The assay uses the hydrophobe fluorescent probe such as N-phenyl-1-naphthylamine (NPN) that incorporates in hydrophobe regions and/or interacts with hydrophobe interfaces [19]. As amphiphiles, polysorbate forms micelles above a specific concentration. As a consequence of the physico-chemical properties of NPN, the dye incorporates in the hydrophobic core of the micelles. With increasing micelle concentration, the fluorescent signal increases and allows a direct determination of the polysorbate concentration, more precisely the presence of micelles [28]. The advantages of this method are its simplicity and assay rapidness. The major drawback is that the method only detects polysorbate compounds that form micelles and it does not differentiate between different polysorbate components [29], [30]. Furthermore, as polysorbate degrades, FFA are released and also might form micelles.

The progress in the development of analytical instruments has resulted in a number of polysorbate methods being based on reversed-phase high performance liquid chromatography (RP-HPLC or UHPLC) coupled with a variety of different detectors. This enabled a more robust detection of several different polysorbate components. Detection methods that are based on evaporation techniques like evaporation light scattering (ELSD) and charged aerosol (CAD) are widely used [30], [31]. To separate the analyte from the protein, mostly mixed mode columns are applied, at which the polysorbate components are bound. The chromatographic separation of the polysorbate components is carried out on a second column [32]. Depending on the protein format used in the formulation, it is sometimes challenging to separate the protein completely and to avoid a remaining concentration of protein in addition to polysorbate on the column (protein carry-over). This results in distorted peaks which go along with an inaccurate determination of polysorbate. Nevertheless, these methods are helpful to support the qualitative determination, the polysorbate finger print, and monitor differences in polysorbate components by overlaying several chromatograms of different samples, for instance. For a more precise determination of polysorbate composition, mass spectrometry (MS) is widely used [4], [5], [21], [33] and even hydrodynamic profiling (HAP) by NMR spectroscopy was recently used for polysorbate analysis [34]. Only a few methods are available to quantify specific components of polysorbate by MS in biopharmaceutical formulations. However, these methods use complex techniques like coupling of two-dimensional liquid chromatography with MS [32] or monitor just selected polysorbate compounds as the esterified and ethoxylated species [13], [15].

The reason of the presented new method development was to set up a stability-indicating method for the main components of polysorbate 20 and their hydrolytic degradation products. Therefore, the aim of the present study was to develop an assay for the quantification of a large number of polysorbate 20 compounds, obtained by chemical hydrolysis, using just one method without the need of derivatization and complex detection techniques. Furthermore, it allows to select specific marker peaks that assesses the potential degradation pathway. Based on this consideration, the QDa detector, a single quadrupole mass spectrometer, appears to be very suitable for this purpose. This detection technique was already used for the quantification of free fatty acids during polysorbate hydrolysis [35], [36], but was not used for the monitoring and selective marker-based quantification of different polysorbate components so far. The QDa mass detector combines the analytical confidence and sensitivity of a mass detection. In addition, the risk of unexpected co-elutions of components is minimized. Its reduced complexity of the tuning options and its generic detection parameters leads to a straightforward use of the detector [35].

This paper describes the method development and validation to reliably quantify marker-based components of polysorbate 20.