RU2805603C2 - Method of quantitative determination of polysorbates in a sample including lc-ms with internal standard - Google Patents

Abstract

FIELD: analytical chemistry.

SUBSTANCE: following is disclosed: a method of quantifying at least one polysorbate or polysorbate oxidation and/or hydrolysis product in a sample, comprising the step of performing an LC-MS analysis of the said sample based on a dioxolanilium ion signal and the step of performing an internal calibration using an internal standard of the said polysorbate. Also a method of monitoring the degradation of at least one polysorbate in a sample is disclosed.

EFFECT: group of inventions provides an effective and reliable method that allows to perform the quantitative determination of polysorbates in a single analysis.

13 cl, 10 dwg, 2 tbl, 1 ex

Description

The present invention relates to analytical methods for the quantitative determination of polysorbates in biotherapeutic samples. Polysorbates, such as PS80, are the most commonly used surfactants in biotherapeutics.

Polysorbates are amphiphilic nonionic surfactants commonly used in biopharmaceutical formulations. Their main role is to protect proteins and monoclonal antibodies (mAbs) from interfacially induced aggregation. Polysorbates such as PS80 are a heterogeneous mixture of more than 1500 molecules covering a wide range of physicochemical properties and including a variety of fatty acids. PS80 is defined as polyoxyethylene sorbitan monooleate, but due to the way it is synthesized, it is a heterogeneous mixture also known to be susceptible to autoxidation and enzymatic hydrolysis of esterified fatty acids. The main source of PS80 heterogeneity can be attributed to the ester linkage and several fatty acids involved in it. Indeed, several studies have reported the presence of large amounts (up to 34.1%) of diesters and sometimes higher order esters (Borisov et al. Pharm. Biotechnol. 2015, 104, 1005-1018). This complexity poses a huge challenge for PS80 quantification and degradation monitoring.

A recent review (Martos et al. J. Pharm. Sci. 2017,106, 1722-1735) describes various methods for monitoring PS80. The presented methods are based on various detection technologies, mainly using liquid chromatography (LC). Among the methods mentioned, those that quantify PS80 are not sensitive to all types of PS80 degradation. For example, methods based on mixed-mode liquid chromatography with ELSD or CAD are suitable for quantifying PS80, but are not the most sensitive methods for PS80 degradation. Mass spectrometry (MS)-based methods are often used to characterize PS80, but not to quantify and monitor PS80 in a quality control (QC) environment. Recent advances in liquid chromatography-mass spectrometry (LC-MS) for PS80 characterization are based on the characteristic signal of the dioxolanilium ion (Borisov et al. Anal. Chem. 2011, 83 , 3934-3942). Reports of the use of these characteristic signals mention their usefulness for the interpretation and identification of PS80 species and especially fatty acid composition. It was also used to elucidate the structure of PS80 oxidation by-products (Borisov et al. 2015, see above). Until now, these signals have not been used to quantify PS80. From the resulting chromatograms, only relative or semi-quantitative information was obtained (Borisov et al. 2011, see above).

Therefore, it remains desirable to provide an efficient and reliable method that allows the quantification of polysorbates in a single assay.

According to one object, the present invention provides a method for the quantitative determination of at least one polysorbate derivative in a sample, the method comprising:

- the stage of performing LC-MS analysis of the specified sample based on the signal of the dioxolanilium ion;

- the step of performing internal calibration using the internal standard of the specified polysorbate.

Although the use of internal standard (IS) is a well-known practice in chromatography and mass spectrometry, the use of IS for the quantification of polysorbates has never been reported. By using an internal standard (IS), linearity, repeatability, accuracy and hence quantification were improved.

One of the key features of the method is the use of an internal standard for internal calibration to avoid problems with matrix effects. Typically, this internal standard has a similar chemical structure and undergoes the same fragmentation at source as the intact polysorbate surrogate. According to the present invention, the method is based on the use of characteristic fatty acid ester dioxolanilium ions resulting from dissociation in the source.

Through the use of a carefully selected internal standard, this method allows the quantification of polysorbates in a typical mAb formulation.

According to one embodiment, the polysorbates are a mixture of compounds with formula (I):

(I)

where w, x, y and z, whether the same or different, independently represent the number of units of polyethylene glycol in formula (I) and differ from 0.

According to one embodiment, w+x+y+z=20, where the polysorbate is, in particular, PS20 or PS80.

In one embodiment, said polysorbate is PS20 or PS80. PS20 and PS80 are commercially available (particularly from Seppic-Puteaux, France).

In one embodiment, said LC-MS analysis uses a single quadrupole mass detector such as QDa® (available from Waters). By applying in-source fragmentation, dioxolanilium ions are generated from polysorbate in the QDa mass detector. QDa can perform multiple single ion detection (SIR) signals in parallel in both positive and negative ionization modes in a single analytical run. This allows for the quantification of intact polysorbate, the search for characteristic byproducts of polysorbate oxidation, and the quantification of free oleic acid, a byproduct of polysorbate hydrolysis. Typically, in the case of PS80, at least four SIR signals were monitored: one for the dioxolanilium oleate ester ion, a specific ion of intact PS80, one for the dioxolanilium IS ion, one for free oleic acid in negative ionization mode (the main by-products of PS80 hydrolysis), one for dioxolanilium ion oxidized oleate ester (major by-products of oxidized PS80). Other signals may be recorded for informational purposes.

In one embodiment, the method monitors the SIR signals of PS80 degradation byproducts: oleic acid (negative m/z 281.3) in the case of hydrolysis and oxidized oleate ester (positive SIR m/z 325.3) in the case oxidation.

This new method provides at least the same level of information in just one analysis, reducing the time spent analyzing PS80. The information collected this way is much easier to understand since the analytes are indeed specific for either intact PS80 or a marker of degraded PS80.

Therefore, according to one embodiment, the method also includes the step of detecting oxidation and/or hydrolysis of the specified polysorbate. Unlike existing methods, the present method allows the quantification of PS80 in a formulation regardless of whether degradation (oxidation or hydrolysis) or adsorption has occurred, which can reduce the overall concentration of PS80 compounds.

Figure 2(b) illustrates the SIR signals at m/z 309.3 for quantification of intact PS80. In one embodiment, the method allows for the quantification of intact PS80 using a suitable surrogate that is sensitive to both PS80 oxidation and PS80 hydrolysis. Monoesters are more sensitive to degradation and can be used as a surrogate for the quantification of polysorbates. In one embodiment, said specific surrogate for intact polysorbates may be monoesters. For PS80, the surrogate is the compound designated peak 2.1 in Figure 2(b). Typically, depending on the experimental conditions, this peak has a retention time of 6 to 7 minutes; more specifically, under the experimental conditions described below in the experimental section, the peak retention time ranged from 6.85 to 6.95, more specifically about 6.9 minutes. When using this surrogate, the use of this method is relevant for the quantitative determination of intact PS80 in the event of degradation due to hydrolysis or oxidation. None of the previously published methods can do this.

As used herein, the term “internal standard” refers to a chemical compound that is added in a constant amount to samples, negative controls, and calibration standards in chemical analysis. This compound can then be used for calibration by plotting the ratio of the polysorbate signal to the internal standard signal as a function of the concentration of polysorbate in the standards. An internal standard is a compound that is typically very similar, but not identical, to the polysorbate in samples because the effects of sample preparation should be the same, depending on the amount of each compound, for both the signal from the internal standard and the signal(s) ) from the compound of interest in the ideal case. The internal standard used should generally provide a signal that is broadly similar to the polysorbate signal, but different enough that the instrument can readily distinguish between the two signals.

The internal standard should generally be similar in structure to the analyte of interest, have a similar retention time to the analyte of interest, and should undergo similar fragmentation in the source. It must also be stable and not interfere with sample components.

In one embodiment, the internal standard (IS) may be selected from compounds having a polyethylene glycol chain esterified with one or more carboxylic acids to have the same physicochemical properties as the polysorbate. Typically, the carboxylic acid portion must be different from the fatty acids present in the polysorbate mixture.

In one embodiment, the internal standard is an ethoxylated fatty acid with formula (I):

(I)

Where:

is a fatty acid residue

where R represents C3-C24 linear or branched saturated alkyl;

R' is H or where R'' represents a C3-C24 linear or branched saturated alkyl; And

n is defined as the number of PEG (PEG, polyethylene glycol) units in the general formula (I), it is understood that the compound of formula (I) is in the form of a mixture of one or more compounds of formula (I), where n, the same or different, are in the range from 1 to 100;

and mixtures of the above.

In one embodiment, n is an integer from 1 to 100.

According to one embodiment, said internal standard is selected from:

poly(ethylene glycol)bis(2-ethylhexanoate)

where n, defined as above, is from 1 to 100.

More specifically, the internal standard is polyethylene glycol monolaurate (hereinafter referred to as PEG-C12) or polyethylene glycol monomyristate (hereinafter referred to as PEG-C14); even more specific:

(polyethylene glycol myristate, PEG-C14).

PEG-C12 and poly(ethylene glycol) bis(2-ethylhexanoate) are commercially available from Sigma-Aldrich (Saint Quentin Fallavier, France), and PEG-C14 is used as a component present in PEG-C12.

In the case of PEG-C12, n may be such that the average molecular weight is from 300 to 600 g/mol, typically about 400 g/mol.

In the case of poly(ethylene glycol) bis(2-ethylhexanoate), n may be such that the average molecular weight is from 400 to 700 g/mol, typically about 650 g/mol.

PEG-C14 can be isolated from PEG-C12 or can be used through PEG-C12 as a component present in PEG-C12.

It has been found that the detection of polysorbate monoester and higher order ester and free oleic acid, as well as the sensitivity to polysorbate and free oleic acid, can be adjusted by adjusting the mobile phase and/or cone voltage.

In one embodiment, the mobile phase of the LC-MS phase contains a formic acid gradient. Typically the mobile phase is a ternary mobile phase containing water, an organic solvent such as acetonitrile or isopropanol, and an acid such as formic acid or acetic acid. Typically the mobile phase consists of water, acetonitrile and formic acid with a gradient of acetonitrile and formic acid, usually with a formic acid gradient ranging from 0.01 to 0.1%. Typically, the composition of the elution phase varies from an initial composition containing 80-90% water, 10-20% acetonitrile and 0.01-0.05% formic acid, to a final composition containing 0-10% water, 90-99.9% acetonitrile and 0.05%-0.1% formic acid (in parts).

Cone stress can cause fragmentation at the source. Typically, the cone voltage is below 100 V, preferably below 70 V. Typically, different cone voltage (CV) can be used for polysorbate and oleic acid. As an illustration, a CV of 50 for PS80 and a CV of 15 for oleic acid can be used to achieve higher sensitivity for PS80 and free oleic acid.

According to one embodiment, the method includes the initial step of adding an internal standard to the sample. The method may also include the step of treating the sample to precipitate proteins that it may contain. Said treatment may include adding acetonitrile to the sample and/or centrifuging said sample.

According to one embodiment, the sample is a biopharmaceutical composition.

In this context, a biopharmaceutical composition is a pharmaceutical composition containing at least one biological product, such as a protein. Accordingly, said sample may contain at least one therapeutic protein, biotherapeutic or biological drug, such as a monoclonal antibody (mAb) or fragment thereof, single chain variable fragment (scFv), single domain antibody (VHH antibody, i.e. nanobody), antibody-drug conjugate, bispecific antibody, trispecific antibody.

According to another object, the present invention also relates to a method for monitoring the degradation of at least one polysorbate in a sample. Said method involves implementing a method for the quantitative determination of said polysorbates according to the invention.

Figures

Figure 1 presents the theoretical structure and degradation pathways of PS80.

Figure 2 shows the dependence of intensity (in arbitrary units) on retention time (in minutes). Illustrated is the characteristic total ion current (TIC) profile of PS80 with peaks labeled 1 - non-esterified species, 2 - monoesters and 3 - higher order esters (a), single ion record (SIR) for m/z 309.3. indicating elution of oleate compounds with peaks labeled 2.1 - POE sorbitan monooleate, 2.2 - POE isosorbide monooleate and 2.3 - POE monooleate (b) and SIR for m/z 283.3, indicating elution of palmitate species (c).

Figure 3 shows chromatograms of dioxolanilium oleate ester ions in positive ionization mode with different percentages of formic acid with a CV of 50 (a) and with different CVs at 0.1% formic acid (b). The peak area extracted from (b) is plotted against cone voltage for the monoester and the higher order ester (c).

Figure 4 shows chromatograms of oleic acid in negative ionization mode with different percentages of formic acid with a CV of 15 (a) and with different CVs at 0.01% formic acid (b).

Figure 5 shows typical chromatograms for the differential diagnosis of PS80 degradation in thermally stressed (upper chromatogram) and unstressed (lower chromatogram) samples. PS80 oleate subspecies are reduced in the temperature-stressed sample (a - positive m/z 309.3). There was no increase in oleic acid content (b - negative m/z 281.3), but the amount of oxidized by-products increased sharply (c - positive m/z 325.3).

Figure 6 shows typical differential diagnostic chromatograms for PS80 degradation in two different batches A (upper chromatogram) or B (lower chromatogram). PS80 oleate subspecies are reduced in the batch A sample (a - positive m/z 309.3). There was a significant increase in oleic acid content (b - negative m/z 281.3), and oxidized by-products did not differ between the two batches (c - positive m/z 325.3).

Figure 7 shows the SIR signals in positive ionization mode at 309.3 m/z for sample PS80 at 50 μg/ml (dashed line) and internal standard (solid line) for PEG bis C8 (a) and PEG C12 and PEG C14 ( b).

Figure 8 is an enhanced ion chromatogram in positive ionization mode at 171 m/z specific PEG bis C8 signal for sample PS80 (dashed line - panel b) and PEG bis C8 (solid line - panel a). The retention time (Rt) of the peak used as an internal standard is given.

Figure 9 is an extracted ion chromatogram in positive ionization mode at 227.3 m/z specific signal PEG C12 for sample PS80 (dashed line - panel b) and PEG bis C8 (solid line - panel a). The retention time (Rt) of the peak used as an internal standard is given.

Figure 10 shows the single ion recording in positive ionization mode at the specific signal of 255.3 m/z PEG C14 for sample PS80 (dashed line) and PEG C14 (solid line). The retention time (Rt) of the peak used as an internal standard is given.

Examples

Materials. Polysorbate 80 was obtained from Seppic (Puteaux, France). For LC-MS, 1 g of polysorbate was dissolved in 100 ml of water in a volumetric flask to obtain a 10 g/L stock solution, stored at 2-8 °C, protected from light. LC-MS grade acetonitrile was purchased from Fisher Scientific (Illkirch, France). Purified water from the milliQ system was used. Formic acid, polyethylene glycol monolaurate (PEG-C12, Mn=±400 g/mol), polyethylene glycol bis-2-ethylhexanoate (PEG-bis C8, Mn=±650 g/mol) and oleic acid were purchased from Sigma-Aldrich ( Saint Quentin Fallavier, France). PEG-C12 was used as an internal standard. A stock solution of PEG-C12 was prepared by dissolving 500 mg in 100 ml acetonitrile in a volumetric flask to obtain a stock solution of 5 g/L. A working solution of 100 μg/ml was prepared by dilution in acetonitrile. Only one batch of PEG-C12 was used throughout the study. A stock solution of oleic acid was prepared by dissolving 100 mg in 100 ml acetonitrile in a volumetric flask to obtain a stock solution of 1 g/L. A working standard solution for calibration was prepared in a 20 mL volumetric flask with a final concentration of PEG-C12 of 5 μg/mL, varying the concentration of PS80 from 5 to 75 μg/mL, and varying the concentration of oleic acid from 1 to 20 μg/mL. The added solvent consisted of a mixture of water/acetonitrile (20%/80%).

LC-MS analysis. Reversed phase separation was performed on an Acquity UPLC system equipped with a QDa mass detector from Waters (Saint Quentin en Yvelines, France). The QDa parameters were set to default prior to the optimization detailed below. A Zorbax Sb-Aq column (100 × 2.1 mm, 3.5 μm) from Agilent (Les Ulis, France) was operated at 50 °C with a flow rate of 1 ml/min according to Christiansen et al. ( Pharmazi. 2011, 66 , 666-671). Mobile phases A and B were water + 0.1% formic acid and acetonitrile + 0.1% formic acid, respectively. The analytical gradient was as follows: 85% A with 15% B for 1 minute followed by a ramp to 60% B after 6 minutes, hold for up to 8 minutes before ramping to 100% B after 10 minutes, hold for another 3 minutes and then return to the initial conditions after 13.1 minutes for another 3 minutes. A typical mobile phase gradient is described in the following gradient table:

Table 1. Mobile phase gradient graph

A (Water) B (acetonitrile) C (acetonitrile + 0.1% formic acid) original 85% 5% 10% 1 min 85% 5% 10% 6 min 40% 50% 10% 8 min 40% 50% 10% 9 min 20% 70% 10% 9.1 min 18% 0% 82% 10 min 0% 0% 100% 13 min 0% 0% 100% 13.1 min 85% 5% 10%

Analytical conditions such as mobile phase and gradient were adjusted according to the results obtained.

Preparation of samples. Samples of a biotherapeutic drug containing mAbs and various excipients were used. The concentration of PS80 was maintained at 200 μg/ml. Sample preparation consisted of the step of protein precipitation with acetonitrile. To 80 μl of the prepared mAb, 20 μl of internal standard working solution and 300 μl of acetonitrile were added. After stirring, the mixture was centrifuged for 10 min at 1500 g at 10°C. The supernatant was collected and transferred to HPLC vials. The final concentration was approximately 40 μg/mL PS80 and 5 μg/mL internal standard.

Results and discussion

Separation of PS80 components and identification of fatty acid esters with in-source dissociation. According to a large number of previous publications, polysorbates are a heterogeneous mixture with great diversity in chemical structure and concentration of esterified compounds. Moreover, the nature of the fatty acids involved in the ester linkage may vary. According to the EU Pharmacopoeia, oleic acid and palmitic acid (monounsaturated and saturated fatty acids, respectively) are the two main PS80 fatty acids. A series of low m/z ions characteristic of polyethylene glycol fatty acid esters, called dioxolanilium ions, were used to identify esterified PS80 species. A typical PS80 profile shows a variety of fatty acid esters from mono- to tetraesters (Figure 2a), while single ion detection (SIR) m/z 309.3 and 283.3 are specific to the oleate ester and palmitate ester species ( figure 2b and 2c). Chromatograms obtained with SIR were much easier to interpret and integrate due to the artificial increase in chromatographic resolution. The identification of the different ester species was based on the work published by Borisov et al., 2015 (see above) on the elution order of polysorbates in reverse phase chromatography (Figures 2a and 2b).

Development of the LC-MS method. QDa can perform simultaneous negative and positive ionization analysis for an ESI source. To take advantage of this capability, LC-MS conditions were carefully optimized. Systematic evaluation of the percentage of formic acid in the mobile phase and the cone voltage (CV) of the ESI source allowed simultaneous separation of PS80 and its degradation product with high sensitivity. Formic acid percentages between 0.01% and 0.1% had little effect on detection of oleate monoesters but were critical for higher order esters (Figure 3a). Increased cone voltage resulted in an increase in the signal represented by the peak area of the monoester and higher order ester (Figures 3b and 3c), up to CV50. Above CV50 there was a sudden loss of signal. This phenomenon was attributed to fragmentation at the source. Increased cone voltage, induced by increased fragmentation at the source, resulted in the formation of more dioxolanilium ions up to CV50. For higher CV, intense fragmentation at the source resulted in the formation of a secondary fragment, accounting for the loss of signal.

To detect oleic acid in negative ionization mode, a small amount of formic acid in the mobile phase was required - no more than 0.01% (Figure 4a). Detection of the unfragmented oleic acid molecular ion required a low cone voltage to minimize fragmentation at the source (Figure 4b).

The elution order between monoesters, oleic acid, and higher order esters allowed the establishment of a three-component mobile phase gradient to vary the percentages of acetonitrile and formic acid independently throughout the analysis. The LC-MS method was adapted from the experimental section as follows. Mobile phase gradient with 85% A (water), 5% B (acetonitrile) and 10% C (acetonitrile + 0.1% formic acid) for 1 minute followed by a linear increase to 40% A, 50% B and 10% C from 6 minutes was held for 2 minutes. Another rise was added to 20% A, 70% B and 10% C in 9 minutes. At 9.1 minutes a switch to 18% A with 82% C was added to increase the percentage of formic acid to 0.1% and the ramp continued to 100% C at 10 minutes and held for 3 minutes before returning to the initial conditions at 13.1 minutes. A switch between mobile phase B and C after 9 min was added to change the percentage of formic acid in the mobile phase from 0.01% to 0.1% while maintaining the percentage of acetonitrile in the mobile phase at the same gradient level. SIR signals in positive ionization mode were recorded using CV50, while SIR signals in negative ionization mode were recorded using CV15.

Quantitative determination of polysorbate using external calibration. In previously published methods, PS80 or PS20 is quantified using a suitable surrogate with an external calibration curve. Well-known examples are processes based on hydrolysis and subsequent esterification in methanol to form a fatty acid methyl ester such as methyl oleate. This methyl oleate is then considered a surrogate for PS80 and quantified. In the present example, the surrogate for quantification was peak 2.1 (Figure 2b), since this peak corresponds to POE sorbitan monooleate, which is a structure close to the theoretical structure of PS80. Based on this peak, a calibration curve was constructed for seven concentration levels with PS80 concentrations in aqueous solution ranging from 5 to 75 μg/mL. Linearity was assessed using unweighted linear regression of peak area as a function of concentration. Linearity was poor with r²=0.977. The residuals showed quadratic behavior, explaining the poor linearity performance. Reproducibility, assessed by introducing 6 replicates of each calibration level, was poor with RSD% peak area above 22%. It was hypothesized that these poor results were due to competition during in-source fragmentation, since this fragmentation is poorly controlled in the ESI QDa source, a single quadrupole.

Comparison of external and internal calibration. To overcome this problem, an internal standard (IS) was used to perform internal calibration instead of external calibration. The most commonly used internal standard in LC-MS for quantitation is deuterated compounds. Given the heterogeneity of the PS80 mixture, this approach was not considered. Compounds with similar chemical structures and therefore similar fragmentation pathways in the sources were selected: polyethylene glycol monomyristate (PEG-C14), polyethylene glycol monolaurate (PEG-C12) and polyethylene glycol bis-2-ethylhexanoate (PEG-bis C8). PEG-C12 and PEG-bis-C8 are commercially available (Sigma-Aldrich/Merck). PEG-C14 could not be purchased from classical chemical manufacturers, but was discovered as an impurity of PEG-C12.

Because this PEG-C14 consists of PEG esterified with C14 saturated fatty acids, it underwent the same fragmentation, resulting in the formation of characteristic dioxolanilium ions at m/z 255.3 in positive ionization mode. The interaction between PS80 and PEG-C14/PEG-C12/PEG bis C8 was tested before using it as an internal standard. None of the signals from the PS80 interfered with the PEG-C14/PEG-C12/PEG bis C8 signal and vice versa. The surrogate peak retention times of PEG-C14/PEG-C12/PEG bis C8 and PS80 were similar, such that they were subject to the same competition during fragmentation at the source.

The lack of interference of signals from PEG-C14/PEG-C12/PEG bis C8 with PS80 signals is shown in Figure 7.

The lack of interference of signals from PS80 with signals from PEG-C14/PEG-C12/PEG bis C8, as well as the retention time of the peak of interest from PEG-C14/PEG-C12/PEG bis C8, is illustrated in Figures 8-10.

For comparison, internal calibration was assessed using the same set of experiments as external calibration. Linearity was assessed using an unweighted linear regression of the ratio of PS80 surrogate peak areas to IS as a function of concentration. Linearity was good with r ² greater than 0.999. The residuals were distributed more or less randomly, and the relative systematic error did not exceed 10%. Reproducibility, assessed by running 6 replicates of each calibration level, was also good with peak area ratio RSD% below 6% for all calibration levels. Table 2 compares these reproducibility results for both external and internal calibration using polyethylene glycol monomyristate (PEG-C14), polyethylene glycol monolaurate (PEG-C12), polyethylene glycol bis 2 ethylhexanoate (PEG-bis C8). Comparison of these ISs is based on linearity (r ² ) and repeatability (RSD%).

Table 2. RSD% peak area (external calibration) and peak area ratio (internal calibration) for 6 replicates (except PEG bis C8, 4 replicates only)

Linearity Target concentration (µg/ml) r² 5 7.5 10 20 thirty 50 75 External calibration 0.977 27.0 27.8 27.1 24.8 23.4 22.8 23.1 Internal calibration PEG C14 0.999 4.8 5.4 4.3 4.3 4.6 5.1 5.5 PEG C12 0.991 6.1 5.5 4.3 3.0 3.4 4.4 5.9 PEG bis C8 0.988 6.1 3.8 3.5 1.8 2.2 1.7 1.3

Based on these data, internal calibrations such as PEG-C12, PEG-C14 and PEG bis C8 were shown to exhibit higher repeatability compared to external calibration. These internal standards demonstrate similar repeatability.

Quantitative determination of oleic acid. Oleic acid is the main by-product of PS80 hydrolysis. Several methods described in Martos et al. (2017, supra), were designed to quantify free fatty acids, including oleic acid, released upon hydrolysis of PS20 or PS80.

Due to the features of QDa, the SIR signal of free oleic acid (m/z 281.3 in negative ionization mode) was recorded in one analytical run along the SIR signals of intact and oxidized PS80. Oleic acid was determined by external calibration with six calibration levels - from 1 µg/ml to 20 µg/ml - added to the PS80 calibration level. For example, a typical calibration level with 20 μg/mL PS80 also contained 5 μg/mL internal standard and 5 μg/mL oleic acid. Linearity was assessed using unweighted linear regression of peak area as a function of concentration. Linearity was good with r ² greater than 0.997 and random distribution of residuals. Reproducibility was assessed similarly to PS80. RSD% values ranged from 6.0 to 13.1% with the maximum value observed at a concentration of 2 μg/ml. Although the RSD% values were higher than expected, the analysis was considered satisfactory as this information would only be used if the study were to be conducted. It was suggested that these high values were caused by high baseline noise in the negative ionization mode due to the composition of the mobile phase. Indeed, a compromise was made during development to favor the detection of PS80 oleate ester species.

Application to prepared mAb samples. The method was applied to different cases of mAb formulations with different mAb properties and excipients, but each time with 200 μg/ml PS80.

mAb1 at a concentration of 5 mg/ml was used to evaluate the repeatability of sample preparation as well as the accuracy of the method. Three preparations were made and analyzed over three consecutive days, resulting in nine PS80 concentrations. The overall mean was 182 μg/mL with an RSD of 3.1%. These results were consistent with previous measurements with the classic mixed-mode LC-CAD assay, where a mean value of 188 μg/mL was found with a 1.2% RSD across six measurements.

mAb2 at 20 mg/ml was exposed to thermal stress for two weeks at 40°C. The results showed a dramatic decrease in PS80 from 198 to less than 25 μg/ml. More information was obtained from other recorded signals. The palmitate ester species chromatogram showed no differences between stressed and unstressed samples, indicating that degradation only affects the oleate ester species typically observed with oxidation. No traces of free oleic acid were found, ruling out hydrolysis as a degradation route. SIR m/z 325.3 signals (in positive ionization mode) showed large differences between stressed and unstressed samples (Figure 5). Ions at m/z 325.3 were described as characteristic by-products of the oxidation of PS80 as polysorbate modified with epoxy stearate or hydroxyl oleate. Here, a complete picture of PS80 in this degraded sample was obtained using three different chromatograms recorded in one analysis, which allows us to clearly conclude that the decrease in PS80 was due to oxidation.

Another example is mAb3 at 50 mg/ml. Samples from different batches were stored at 5°C for almost a month before analysis. The results showed differences in PS80 content between batches A and B, respectively 100 and 190 μg/ml instead of 200 μg/ml. This is illustrated in Figure 6a when considering peak 2.1, a surrogate for PS80 quantification. The SIR signals of oxidized PS80 showed no difference between batches, meaning there was no increase in PS80 oxidation byproducts, especially compared to the intensity observed with mAb2. Free oleic acid in solution was detected in problem batch A at a concentration of approximately 25 µg/ml. In batch B, only small traces of oleic acid were detected (Figure 6). The presence of free oleic acid indicates hydrolysis of PS80. Considering all the information obtained from this analysis, it was concluded that the degradation of PS80 occurs due to hydrolysis. Since higher order esters were not affected, this hydrolysis was assumed to be of enzymatic origin.

Conclusion. PS80 is widely used in biotherapeutics. Recently, there has been growing concern about the stability of polysorbates in drug formulation and their ability to maintain their protective role against protein aggregation. Effective methods for measuring PS80 and monitoring its degradation have been reported over the past few years (Martos et al., supra). Among them, LC-MS based methods have shown promising results in terms of characterization and semi-quantitative information. The characteristic fatty acid ester dioxolanilium ion signals were used after in-source CID to greatly simplify the chromatograms and more easily identify PS80 subclasses. In the previous work mentioned, the method was not tested in a quantitative analysis. The present study applied the same methodology with the aim of using it for quality control level quantification using a single quadrupole mass detector. Unfortunately, the first results of the quantitative assessment confirmed Borisov's conclusion (see above). This problem was resolved by using a carefully selected internal standard with similar chemical properties. A formic acid gradient has been shown to allow greater sensitivity and simultaneous detection in both positive and negative ionization modes of all compounds of interest from PS80 ester subclasses to free fatty acids.

This method was successfully applied to various PS80 monitoring cases and especially to two of them with distinctive PS80 degradation features. In these two cases, the root cause of PS80 degradation was identified using only one 16-minute analysis. To achieve the same goal without this method would require three to four different methods: one to quantify PS80 (mixed-mode LC-CAD) plus one to profile PS80 (reverse-phase CAD), as well as other methods to determine characteristic by-products (oxidized PS80 and free oleic acid). The method presented herein can effectively replace the extensive set of analytical tools that have been required hitherto.

The purpose of this method is to provide an accurate and stable measurement of PS80, as well as provide valuable information to unambiguously determine the root cause of PS80 degradation.

Claims (26)

1. A method for the quantitative determination of at least one polysorbate or a product of oxidation and/or hydrolysis of polysorbate in a sample, comprising: - the stage of performing LC-MS analysis of the specified sample based on the signal of the dioxolanilium ion; - the step of performing internal calibration using the internal standard of the specified polysorbate, where the internal standard is an ethoxylated fatty acid with formula (I): (I) Where: represents a fatty acid residue, where R represents C ₃ -C ₂₄ linear or branched saturated alkyl; R' is H or where R'' represents C ₃ -C ₂₄ linear or branched saturated alkyl; And n ranges from 1 to 100; or mixtures thereof. 2. The method according to claim 1, wherein said polysorbate is PS20 or PS80. 3. A method according to any one of the preceding claims, wherein said internal standard is selected from: where n is from 1 to 100. 4. The method according to any of the preceding claims, wherein said internal standard is , and n is from 1 to 100. 5. The method of any one of the preceding claims, wherein said LC-MS analysis uses a single quadrupole mass detector (QDa). 6. A method according to any one of the preceding claims, wherein the LC-MS mobile phase contains a gradient of formic acid. 7. The method according to any one of the preceding claims, wherein the LC-MS mobile phase is a ternary mobile phase. 8. The method according to claim 7, in which the ternary mobile phase contains water, acetonitrile and formic acid. 9. The method according to any of the preceding claims, wherein said sample is a biopharmaceutical composition. 10. The method according to any of the preceding claims, wherein said sample contains at least one protein. 11. The method according to any of the preceding claims, wherein said sample contains at least one monoclonal antibody. 12. A method according to any one of the preceding claims, wherein said method also includes the step of detecting oxidation and/or hydrolysis of said polysorbate. 13. A method for monitoring the degradation of at least one polysorbate in a sample, comprising implementing the method of any one of the preceding claims.

Images