Extremely Simple and Rapid HPLC Analysis of Tocilizumab in Human Serum with Selective Precipitation Using Alkylamine

Abstract

An assay using HPLC with fluorescence (FL) detection method for monitoring native FL of tocilizumab (TCZ) in human serum combined with extremely simple and rapid pretreatment without any antigen-antibody reaction was developed. Good separation of TCZ was achieved within 13 min on a Presto FF-C18 column (100 × 4.6 mm i.d., 2 µm). Simple pretreatment with acetonitrile containing primary and secondary alkylamines having longer than C3 in the alkyl chain removed immunoglobulin G subclass 1 and TCZ could be recovered selectively. The spiked calibration curve of TCZ in human serum showed good linearity in the range of 40–1000 µg/mL (r > 0.997). The lower limit of quantitation (S/N = 10) of the TCZ was 19.7 µg/mL. The accuracy was within 103.5–114.9%, and the intra- and inter-day precisions as relative standard deviations were less than 5.3 and 7.8% (n = 5), respectively. The recovery of TCZ was 42.2 ± 3.4% (n = 3). The TCZ in pretreated sample was confirmed to be stable for 6 h (>95%) at room temperature and 24 h (>95%) at 4 °C. The proposed method is considered extremely superior to the previous methods in terms of time requirement for analysis. Therefore, the developed method may be more useful than conventional methods in urgent situations, such as confirming therapeutic efficacy of cytokine-release syndrome by 2019 coronavirus disease.

Introduction

Therapeutic monoclonal antibodies (mAbs) have been used for over 20 years to treat various diseases and combine high specificity with generally low toxicity.¹⁾ Recently, the increasing importance of therapeutic mAbs is apparent, as mAbs have become a mainstream therapy for a variety of diseases.²⁾ On the other hand, the pharmacokinetic (PK) properties of mAbs differ markedly from those of low-molecular-weight pharmaceuticals, and the clarification of mAbs PK can have important clinical implications.³⁾ Therefore, it is necessary to analyze therapeutic mAbs in biological samples. Ligand-binding assay (LBA), such as the enzyme-linked immunosorbent assay (ELISA), have been general method for analysis of therapeutic mAbs in biological samples.^4–7) The LBA is a highly sensitive and selective analytical method, but has problems such as the risk of cross-reactivity, interference from matrix components, and long development times.⁸⁾ Recently, various liquid chromatography-tandem mass spectrometry (LC-MS/MS) methods as an alternative to LBA have been applied to determine the mAbs in serum or plasma samples.^9–12) Although these methods are highly sensitive and do not require antibody construction like LBA, they have limitations such as time-consuming trypsin digestion and the associated background noise.¹³⁾ In addition, it has been reported that bioanalysis of therapeutic mAbs by high-temperature reversed-phase liquid chromatography (HTRP-LC) with native fluorescence (FL) detection after isolation from human serum using sepharose coupled with anti-idiotype antibodies and immunoaffinity magnetic purification.^14,15) These method does not require complex pretreatment such as trypsin digestion used in LC-MS/MS methods because the mAbs are analyzed directly. And it does not require fluorescence derivatization because of native FL of mAbs. However, these methods also required time-consuming incubation times similar to ELISA and a step of immuno-affinity purification with the anti-idiotype antibodies.

Tocilizumab (TCZ), which is a humanized monoclonal antibody inhibiting the activity of the inflammatory cytokine interleukin-6 (IL-6), is recommended for use in rheumatoid arthritis (RA).¹⁶⁾ Recently, TCZ has been also used in the treatment of acute respiratory distress syndrome (ARDS) as the most devastating complication induced cytokine-release syndrome (CRS) by the novel 2019 coronavirus disease (COVID-19).^17,18) Because C-reactive protein (CRP) as a marker of IL-6 bioactivity is synthesized through IL-6-dependent hepatic biosynthesis, anti-IL-6 activity of TCZ can be monitored by the measurement of CRP in blood.^19,20) On the other hand, since there is generally a delay of up to 24 h from the administration of the drugs, it is difficult to use CRP as an indicator of the therapeutic effect immediately after administration.²¹⁾ Therefore, direct measurement of TCZ concentrations in serum or plasma would be very clinically meaningful. However, the methods measuring therapeutic mAbs, such as LBA and LC-MS/MS, are not appropriate for use in urgent situations such as CRS by the COVID-19 because of the time required to complete the analysis from pretreatment.

In previous research, we could separate TCZ and human immunoglobulin G subclass 1 (IgG1), antibodies belonging to the same subclass, by a combination of simple pretreatment without any antigen-antibody reaction and HTRP-LC-FL.²²⁾ In this study, we examined pretreatment agent such as primary and secondary alkylamines to improve the recovery of TCZ. And the method could be applied to spiked human serum sample.

Experimental

Chemicals and Reagents

TCZ (ACTEMRA^® 20 mg/mL for Intravenous Infusion) was procured from Chugai Pharmaceutical (Tokyo, Japan). Acetonitrile (HPLC grade), n-octylamine (OA), and trifluoroacetic acid (TFA) were obtained from Wako Pure Chemical Corporation (Osaka, Japan). Human serum (pool) was produced by Cosmo Bio Co., Ltd. (Tokyo, Japan). The organic solvent and pretreatment reagents used were of the highest available purity.

Instruments

The HPLC system consisted of two chromatographic pumps (LC-30AD, Shimadzu, Kyoto, Japan), a degasser (DGU-20A5R, Shimadzu), an autosampler (SIL-30AC, Shimadzu), a system controller (CBM-20A, Shimadzu), a Presto FF-C18 column (100 × 4.6 mm i.d., 2 µm, Imtakt Co., Kyoto, Japan), a column oven (CTO-30A, Shimadzu), an RF-20AXS fluorescence detector (Shimadzu), and an LC workstation (LabSolutions/LC solution version 5.92, Shimadzu). A mixture of 0.1% TFA aq. solution (solvent A) and acetonitrile containing 0.1% TFA (solvent B) was used as the mobile phase and the total flow rate was set at 0.5 mL/min. The gradient elution was programmed as follows: 0–1 min (from 10 to 33% B), 1–7 min (from 33 to 38% B), 7–8 min (from 38 to 90% B), 8–10 min (from 90 to 95% B), and 10–13 min (10% B). The column temperature was set at 75 °C, and eluates were monitored at 288 nm (λex) and 346 nm (λem).

Sample Preparation

Human serum was diluted 2-fold with a saline solution, and then 150 µL acetonitrile containing 400 mM OA was added to 150 µL of the diluted serum. After centrifugation at 5000 × g for 3 min, 50 µL of 7.0% TFA was added to the supernatant (200 µL). Following filtration through a membrane filter (0.2 µm, Minisart^® RC, Sartorius, Göttingen, Germany), a 2.5-µL aliquot of the sample solution was injected into the HPLC system.

Preparation Calibration Standards and Quality Control Samples

TCZ standard solution was serially diluted with human serum to obtain spiked calibration standards at concentrations of 40, 100, 200, 500, and 1000 µg/mL. To assess the accuracy and precision, quality control (QC) samples at 40, 200, and 1000 µg/mL of TCZ in diluted drug-free human serum were prepared.

Methods Validation

Proposed analytical method was (limit of detection (LOD), lower limit of quantitation (LLOQ), linearity, accuracy, intra- and inter-day precisions, and recovery) followed ICH M10 Bioanalytical method validation.²³⁾

The LOD and LLOQ were defined as the concentration at a signal-to-noise ratio of 3 (S/N = 3) and S/N = 10, respectively. The linearity was calculated by the spiked calibration curves. The precision and accuracy of the method were determined by triplicate measurements of three concentrations for each compound. Recovery was calculated as the ratio of the slope of the spiked calibration curve to that of the standard.

Results and Discussion

Separation of TCZ

The separation of TCZ was examined using Phenomenex Aeris Widepore XB-C8, Waters BioResolve RP mAb polyphenyl column, and Imtakt Presto FF-C18 columns. Among them, the Presto FF-C18 column gave the best separation of the TCZ and the lowest symmetry factor (S = 1.8), and the superior peak shapes.

Examination of Alkylamines in Pretreatment

In previous research, we could separate standard TCZ and IgG1, these antibodies belonging to the same subclass, by a combination of simple pretreatment with acetonitrile containing bases and HT-RPLC with the non-porous analytical column to some degree.²²⁾ In particular, TCZ was relatively well recovered when tripropylamine was used as the base. The separation of TCZ and IgG1 which is present in high concentration in blood, is required to establish practical analytical method. The pretreatment method was investigated using human serum by primary, secondary, and tertiary alkylamines as bases in preliminary study, and found that both primary (methylamine, ethylamine, n-propylamine, n-butylamine, and OA) and secondary amines (dimethylamine, diethylamine, di-n-propylamine, and di-n-butylamine) completely removed large amounts of IgG1 in human serum. However, tertiary amines (trimethylamine, triethylamine, and tripropylamine) at any of the concentrations examined could not remove the IgG1-derived contaminant peak on chromatogram. The alkylamines having more longer than C3 in the alkyl chain could be recovered TCZ selectively. These suggest that the removal of IgG1 is related to the hydrophobicity and steric hindrance of the alkylamine used in the pretreatment. In addition, primary and secondary amines which exhibit similar values of logP were examined. As results, the peak area value of TCZ increased with increasing the length of alkyl chain. And recoveries for primary amines were higher than those for secondary amines, (Fig. 1). As result, n-octylamine was selected in the following experiments, because it gave the highest recovery of TCZ. In this study, alkylamines soluble in acetonitrile were used.

Fig. 1. Effects of (a) the Length of Alkyl Chain, and (b) Log P Value on the Recovery of TCZ

Pretreatment Conditions

To increase the recovery, pretreatment conditions such as OA and TFA concentrations were investigated. The OA concentrations in the range of 10–500 mM and 5.5–8.0% for TFA were used. The highest relative peak area of TCZ was obtained when 400 mM of OA and 7.0% TFA was used (Fig. 2). Addition of TFA gave stable statement of sample solution. So, centrifugation time before TFA addition should be controlled stringency. The centrifugation time was examined. Less than 2 min of centrifugation show insufficient separation and more than 10 min decrease peak area of TCZ. So, 3 min of centrifugation was selected in following experiments.

Fig. 2. Effects of (a) OA, and (b) TFA Concentration on the Recovery of TCZ

Figure 3 shows typical chromatograms obtained from (a) serum spiked with 200 µg/mL of TCZ and (b) blank serum. The TCZ peak was acceptably separated from contaminant and other interfering peaks within 13 min, and the retention times of TCZ was 7.9 min. On the other hands, chromatogram obtained from spiked serum pretreated without OA (Fig. 3 (c)) showed no difference with that of blank (Fig. 3 (d)). These results suggest that OA contributes to the selective precipitation.

Fig. 3. Chromatograms of (a) Serum Spiked with 200 µg/mL of TCZ and (b) Blank Serum Pretreated with OA, and (c) Spiked One and (d) Blank Pretreated without OA

The stabilities of the TCZ in pretreated sample in the autosampler maintained at room temperature (r.t.) and 4 °C were determined. As shown in Fig. 4, TCZ was confirmed to be stable for 6 h (>95%) at r.t., and was confirmed to be stable for 24 h (>95%) at 4 °C. Molecular colloids such as TCZ are more stable at lower temperature,²⁴⁾ so TCZ would have been more unstable at r.t. than that at 4 °C in pretreated sample.

Fig. 4. Stability of TCZ on the Pretreated Sample in the Autosampler at Room Temperature and 4 °C

The initial relative peak area was taken as 100.

Method Validation

The spiked calibration curves of TCZ showed good linearities in the ranges of 40–1000 µg/mL (r > 0.997, n = 3). The LOD and the LLOQ of TCZ were 5.9 and 19.7 µg/mL, respectively. These parameters for the proposed method are summarized in Table 1.

Table 1. Calibration Curve, Linearity, LOD, and LLOQ of TCZ

Compound	Calibration range, µg/mL	r*	LOD, µg/mL (S/N = 3)	LLOQ, µg/mL (S/N = 10)
TCZ	40–1000	>0.997	5.9	19.7

* n = 3.

Accuracy, intra- and inter-day precisions, and recovery by the proposed method were evaluated by using serum spiked with known concentrations of TCZ (40, 200, and 1000 µg/mL) and the results are summarized in Table 2. The assay accuracy for TCZ was in the range of 103.5 ± 4.2 and 114.9 ± 14.8%. The intra-day precisions as relative standard deviations (RSDs) and the inter-day precisions (RSDs) were less than 5.3 and 7.8%, respectively. The recovery was calculated by the ratio of the slope of the spiked calibration curve to that of the standard, and those were 42.2 ± 3.4%. Although, the recovery in this study was not so high, the recovery of therapeutic mAbs in the LC-MS/MS method was about 14%.²⁵⁾ The low recovery rate of our method may be attributed to the fact that the insufficient specificity of our pretreatment on TCZ. This means half of TCZ was precipitated with co-existant components. These validation parameters of the proposed method were acceptable for the reliable analyses of the TCZ in the serum sample.

Table 2. Accuracy, Precisions and Recovery of the Proposed Method

Compound	Spiked concentration, µg/mL	Accuracya) %	Precisiona) RSD%		Recoveryb) %
			Intra-day	Inter-day
TCZ	40	114.9 ± 14.8	5.3	7.8
	200	105.4 ± 7.1	4.0	7.0	42.2 ± 3.4
	1000	103.5 ± 4.2	3.6	5.9

a) n = 5. b) n = 3; Ratio of the slope of the spiked calibration curve to that of the standard.

The sensitivity, precision, and time requirement for analysis of the proposed method was compared with those of the previous reports and the results are shown in Table 3. The proposed method is suggested to be less sensitive than LC-MS/MS methods, and be comparable to the LC-MS/MS methods for precision.^26,27) On the other hand, our method is considered to be extremely superior to the LC-MS/MS method in terms of time requirement for analysis. Additionally, the proposed method does not require expensive instruments and reagents for sample pretreatment. The dose of TCZ for CRS with COVID-19 is 8 mg/kg, with a maximum dose of 800 mg, and a second dose is given within 8 to 24 h after the first dose.²⁸⁾ And the maximum concentration (C_max) of TCZ at a single 8 mg/kg dose in adults was reported to be 256 ± 48.3 µg/mL.²⁹⁾ Therefore, this method may be more useful than conventional methods in urgent situations, such as confirming the therapeutic efficacy of CRS by the COVID-19. In this study, we have clarified for the first time that the use of primary and secondary alkylamines with long alkyl chains allows complete precipitation of IgG1 in human serum and recovery of TCZ, which was not possible with previous methods.

Table 3. Comparison of Performances for the Proposed Method with Those of Previous Reports

	Our method	LC-MS/MS
		Chiu et al.26)	Willeman et al.27)
LLOQ, µg/mL	20	13	1
Precision RSD%	<7.8	<12.2	<8.0
Analysis time, h	<0.5	>21	>16

Conclusion

An extremely simple and rapid HPLC analysis assay for the determination of TCZ in human serum with selective precipitation using OA was developed. Well-defined TCZ peaks were obtained by separation, without any interfering peaks, using the proposed method. The obtained validation parameters for TCZ in human serum were acceptable. The sensitivity of the method to confirm the therapeutic efficacy of chronic diseases such as RA was in sufficient, so more detail study to overcome the problem is needed. On the other hand, this method may be more useful than conventional methods in urgent situations, such as confirming therapeutic efficacy of CRS by COVID-19.

Acknowledgments

This work was supported by the JSPS KAKENHI Grant Number 20K16530.

Conflict of Interest

The authors declare no conflict of interest.

References

• 1) Keizer R. J., Huitema A. D., Schellens J. H., Beijnen J. H., Clin. Pharmacokinet., 49, 493–507 (2010).
• 2) Lu R. M., Hwang Y. C., Liu I. J., Lee C. C., Tsai H. Z., Li H. J., Wu H. C., J. Biomed. Sci., 27, 1 (2020).
• 3) Lobo E. D., Hansen R. J., Balthasar J. P., J. Pharm. Sci., 93, 2645–2668 (2004).
• 4) Blasco H., Lalmanach G., Godat E., Maurel M. C., Canepa S., Belghazi M., Paintaud G., Degenne D., Chatelut E., Cartron G., Le Guellec C., J. Immunol. Methods, 325, 127–139 (2007).
• 5) Damen C. W., de Groot E. R., Heij M., Boss D. S., Schellens J. H., Rosing H., Beijnen J. H., Aarden L. A., Anal. Biochem., 391, 114–120 (2009).
• 6) Puszkiel A., Noé G., Boudou-Rouquette P., Cossec C. L., Arrondeau J., Giraud J. S., Thomas-Schoemann A., Alexandre J., Vidal M., Goldwasser F., Blanchet B., J. Pharm. Biomed. Anal., 139, 30–36 (2017).
• 7) Pluim D., Ros W., van Bussel M. T. J., Brandsma D., Beijnen J. H., Schellens J. H. M., J. Pharm. Biomed. Anal., 164, 128–134 (2019).
• 8) Hoofnagle A. N., Wener M. H., J. Immunol. Methods, 347, 3–11 (2009).
• 9) Liebler D. C., Zimmerman L. J., Biochemistry., 52, 3797–3806 (2013).
• 10) Li H., Ortiz R., Tran L. T., Salimi-Moosavi H., Malella J., James C. A., Lee J. W., AAPS J., 15, 337–346 (2013).
• 11) El Amrani M., van den Broek M. P., Göbel C., van Maarseveen E. M., J. Chromatogr. A, 1454, 42–48 (2016).
• 12) Hagman C., Ricke D., Ewert S., Bek S., Falchetto R., Bitsch F., Anal. Chem., 80, 1290–1296 (2008).
• 13) Li H., Ortiz R., Tran L., Hall M., Spahr C., Walker K., Laudemann J., Miller S., Salimi-Moosavi H., Lee J. W., Anal. Chem., 84, 1267–1273 (2012).
• 14) Todoroki K., Nakano T., Eda Y., Ohyama K., Hayashi H., Tsuji D., Min J. Z., Inoue K., Iwamoto N., Kawakami A., Ueki Y., Itoh K., Toyo’oka T., Anal. Chim. Acta, 916, 112–119 (2016).
• 15) Damen C. W., Derissen E. J., Schellens J. H., Rosing H., Beijnen J. H., J. Pharm. Biomed. Anal., 50, 861–866 (2009).
• 16) Farah Z., Ali S., Price-Kuehne F., Mackworth-Young C. G., Biologics., 10, 59–66 (2016).
• 17) Kotch C., Barrett D., Teachey D. T., Expert Rev. Clin. Immunol., 15, 813–822 (2019).
• 18) Chen H., Wang F., Zhang P., Zhang Y., Chen Y., Fan X., Cao X., Liu J., Yang Y., Wang B., Lei B., Gu L., Bai J., Wei L., Zhang R., Zhuang Q., Zhang W., Zhao W., He A., Fr. Medecine, 13, 610–617 (2019).
• 19) Khiali S., Khani E., Entezari-Maleki T., J. Clin. Pharmacol., 60, 1131–1146 (2020).
• 20) Yuasa S., Yamaguchi H., Nakanishi Y., Kawaminami S., Tabata R., Shimizu N., Kohno M., Shimizu T., Miyata J., Nakayama M., Kishi J., Toyoda Y., Nishioka Y., Tani K., J. Med. Invest., 60, 77–90 (2013).
• 21) Hisamuddin E., Hisam A., Wahid S., Raza G., Pak. J. Med. Sci., 31, 527–531 (2015).
• 22) Takada M., Ohba Y., Kamiya S., Kabashima T., Nakashima K., Luminescence, 34, 347–352 (2019).
• 23) International Conference on Harmonization Harmonised Guideline (2019), “Bioanalytical Method Validation and Study Sample Analysis M10.”: ‹https://www.ema.europa.eu/en/documents/scientific-guideline/ich-guideline-m10-bioanalytical-method-validation-step-5_en.pdf›, cited 18 June, 2022.
• 24) Walstra P., J. Dairy Sci., 73, 1965–1979 (1990).
• 25) Heudi O., Barteau S., Zimmer D., Schmidt J., Bill K., Lehmann N., Bauer C., Kretz O., Anal. Chem., 80, 4200–4207 (2008).
• 26) Chiu H. H., Tsai Y. J., Lo C., Liao H. W., Lin C. H., Tang S. C., Kuo C. H., Anal. Chim. Acta, 1189, 339231 (2022).
• 27) Willeman T., Jourdil J. F., Gautier-Veyret E., Bonaz B., Stanke-Labesque F., Anal. Chim. Acta, 1067, 63–70 (2019).
• 28) Hasanin A., Mostafa M., J. Anesth., 35, 896–902 (2021).
• 29) Paccaly A. J., Kovalenko P., Parrino J., Boyapati A., Xu C., van Hoogstraten H., Ishii T., Davis J. D., DiCioccio A. T., J. Clin. Pharmacol., 61, 90–104 (2021).