MS: A preliminary pharmacokinetic study in healthy Chinese volunteers after oral administration of benzonatate soft capsule

Journal of Pharmaceutical and Biomedical Analysis 173 (2019) 134–143

Contents lists available at ScienceDirect

Journal of Pharmaceutical and Biomedical Analysis journal homepage: www.elsevier.com/locate/jpba

Determination of benzonatate and its metabolite in human plasma by HPLC–MS/MS: A preliminary pharmacokinetic study in healthy Chinese volunteers after oral administration of benzonatate soft capsule Jiayu Man a,b,1 , Feifei Jiao b,1 , Yiya Wang b , Yueqing Gu a , Li Ding a,b,∗∗ , Chang Shu a,∗ a b

Department of Pharmaceutical Analysis, China Pharmaceutical University, Nanjing, 210009, China Nanjing Clinical Tech. Laboratories Inc., Nanjing, 211100, China

a r t i c l e

i n f o

Article history: Received 6 November 2018 Received in revised form 15 April 2019 Accepted 15 May 2019 Available online 16 May 2019 Keywords: Benzonatate 4-(Butylamino)benzoic acid Polymer Metabolite HPLC–MS/MS Pharmacokinetics

a b s t r a c t Benzonatate has been used as a non-narcotic oral antitussive drug for many years. Its pharmacokinetics has never been reported due to the technical difficulties in detecting benzonatate by mass spectrometry. However, its concentration can be extrapolated based on the concentration of its metabolite, 4-(butylamino)benzoic acid (BBA). In this study, two sensitive high-performance liquid chromatographytandem mass spectrometry (HPLC–MS/MS) methods were developed and fully validated for the determination of the original 4-(butylamino)benzoic acid (method B) and total 4-(butylamino)benzoic acid (containing the original 4-(butylamino)benzoic acid and 4-(butylamino)benzoic acid converted from benzonatate after collection, method A). For both methods, one-step protein precipitation by methanol was performed to extract analytes from the plasma samples. Chromatographic separation was done on an InfinityLab Poroshell 120 Phenyl Hexyl column (2.1 mm × 50 mm, 2.7 ␮m, Agilent) with initial mobile phase consisting of 5 mM ammonium acetate containing 0.3% formic acid and acetonitrile (60:40, v/v) at a flow rate of 0.3 mL/min. Quantification was achieved by multiple reaction monitoring (MRM) in electron spray ionization (ESI) positive mode with the transitions of m/z 194.2 → 138.1 and 515.3 → 497.3 for 4-(butylamino)benzoic acid and telmisartan (the internal standard), respectively. The two methods exhibited good linearity over the concentration range of 10–10000 ng/mL. Both of the methods were successfully applied to the preliminary pharmacokinetic study in healthy Chinese volunteers after oral administration of benzonatate soft capsule at a single dose of 100 mg. The results showed that 4-(butylamino)benzoic acid and benzonatate were rapidly absorbed and reached a maximum concentration (Cmax ) of 1708 ± 457 ng/mL and 1063 ± 460 ng/mL, respectively. The half-life (t1/2 ) were 1.32 ± 0.29 h for 4-(butylamino)benzoic acid and 1.01 ± 0.41 h for benzonatate. The area under the curve from 0 h to 10 h (AUC0-10 ) for 4-(butylamino)benzoic acid and benzonatate were 2103 ± 918 ng/mL·h and 1097 ± 559 ng/mL·h, respectively. And the data was valuable for further clinical study. © 2019 Elsevier B.V. All rights reserved.

1. Introduction Abbreviations: BBA, 4-butylaminobenzoic acid; HPLC–MS/MS, highperformance liquid chromatography–tandem mass spectrometry; MRM, multiple reaction monitoring; ESI, electron spray ionization; HPSEC, high-performance sizeexclusion chromatography; IS, internal standard; TEM, temperature; CUR, curtain gas; GS1, nebulizer gas; GS2, heater gas; CAD, collision-activated dissociation; DP, declustering potential; CE, collision energy; EP, entrance potential; CXP, collision exit cell potential; LLOQ, lower limit of quantification; QC, quality control; LQC, low QC; MQC, medium QC; HQC, high QC; DQC, dilution QC; Tmax , the time to reach the maximum concentration; Cmax , maximum concentration; t1/2 , half-life; AUC0-10 , the area under the curve from 0 h to 10 h. ∗ Corresponding author. ∗∗ Corresponding author at: Department of Pharmaceutical Analysis, China Pharmaceutical University, Nanjing, 210009, China. https://doi.org/10.1016/j.jpba.2019.05.030 0731-7085/© 2019 Elsevier B.V. All rights reserved.

Cough induced by various causes is the most common symptom in patients, which can be treated by inhibiting different aspects of cough reflex. Benzonatate not only acts peripherally by anesthetizing the vagal stretch receptors in the bronchi, alveoli, and pleura, but also acts upon the central nervous system by inhibiting transmission of impulses of the cough reflex in the vagal nuclei of the

1

E-mail addresses: [email protected] (L. Ding), [email protected] (C. Shu). These authors contributed equally to this work.

J. Man et al. / Journal of Pharmaceutical and Biomedical Analysis 173 (2019) 134–143

medulla, which is widely used for cough treatment in cancer and advanced cancer [1–4]. Moreover, a 7-year retrospective review of all single substance ingestion of benzonatate has been reported by the National Poison Center Database System (NPDS), indicating that the severe medical outcome is uncommon [5]. As a mixture containing 15 components with different degrees of polymerization, there were few studies on the detection of benzonatate. Until now only an HPLC analytical method has been developed to quantify 4-(butylamino)benzoic acid (BBA) and benzonatate in plasma, simultaneously. But it failed to determine the actual concentration of benzonatate in plasma because the lower limit of quantification (LLOQ) of 200 ng/mL was too high [6]. It was well known that the mixture can be usually detected by high-performance size-exclusion chromatography (HPSEC) with refractive index detector [7–11]. However, this method was hardly applied to the detection of trace amount of the mixture in vivo because of low sensitivity. Another, and more importantly, it was difficult to obtain the reference standard of each component of the mixture, which was also a big problem for the direct detection of benzonatate by more sensitive high-performance liquid chromatography-tandem mass spectrometry (HPLC–MS/MS) method. To meet the challenges, an indirect method that the concentration of benzonatate can be estimated based on the concentration of its metabolite (BBA) was proposed. Hence, two sensitive HPLC–MS/MS methods with the LLOQ of 10 ng/mL were developed and fully validated for the determination of the original BBA (method B) and total BBA (containing the original BBA and BBA converted from benzonatate after collection, method A), respectively. It has been reported that the instability of some ester drugs in matrix ex vivo was a common problem, which was mainly caused by the chemical properties, the biological properties of the sample matrix and the storage conditions of the drug-containing samples [12–15]. In order to detect the actual concentration of the original BBA in method B, enzyme inhibitor was added to inhibit the degradation of benzonatate. In method A, drug-containing samples were placed at room temperature for a period to detect the concentration of total BBA. The two validated methods were successfully applied to the pharmacokinetic study of BBA and benzonatate. In this paper, we first published the study to explicitly illustrate the pharmacokinetic profiles of BBA and benzonatate, which provided clinical reference and filled the gap of benzonatate in pharmacokinetic profiles. 2. Materials and methods 2.1. Chemical and reagents Benzonatate was supplied by the United States Pharmacopeia Convention (USA). Tetracaine EP Impurity B (BBA), the metabolite of benzonatate, was purchased from TLC Pharmaceutical Standards Ltd (Ontario, Canada). The internal standard (IS) telmisartan and the enzyme inhibitor neostigmine metilsulfate were obtained from National Institutes for Food and Drug Control (Beijing, China). Sodium fluoride was provided by Sinopharm Chemical Reagent Co., Ltd (Shanghai, China). HPLC grade acetonitrile and methanol were received from Merck KGaA (Darmstadt, Germany). Analytical grade formic acid and ammonium acetate were purchased from SigmaAldrich, Co. (USA). Ultrapure water was freshly produced by a Milli Q water purification system (Millipore, Bedford, MA, USA).

135

API 4000 tandem mass spectrometer (Applied Biosystems/Sciex, Foster, USA). The Shimadzu high performance liquid chromatograph was composed of a system controller (CBM-20A Lite), a binary pump (LC-20AD), a column oven (CTO-20AC), an online degasser (DGU-20A5R) and an autosampler (SIL-30AC). Chromatographic separation was performed on an InfinityLab Poroshell 120 Phenyl Hexyl column (2.1 mm × 50 mm, 2.7 ␮m, Agilent) with a C18 Security Guard Cartridges (2.0 mm × 4.0 mm, Phenomenex) kept at 40 ◦ C. The analyte was eluted by a gradient mobile phase system at a flow rate of 0.3 mL/min. The initial mobile phase was consisted of 5 mM ammonium acetate containing 0.3% formic acid and acetonitrile (60:40, v/v), increased to 90% acetonitrile at 2.3 min and was kept for 0.9 min. Then the mixture was reversed back to 40% acetonitrile at 3.3 min and kept for 1.7 min. Column effluent was veered right into the mass spectrometer by a six-way valve (Applied Biosystems/Sciex, Foster, USA) at the time interval of 1.0–2.2 min, besides the eluent was diverted to the waste. The autosampler temperature was maintained at 8 ◦ C and the sample injection volume was 10 ␮L. ® The mass spectrometer coupled with a Turbo-V ionspray source was operated in the positive ESI mode. Quantification of the analyte and IS was achieved by multiple reaction monitoring (MRM) with the transitions of m/z 194.2→138.1 and 515.3→497.3 for BBA and telmisartan, respectively. The optimized source parameters such as ionspray voltage, temperature (TEM), curtain gas (CUR), nebulizer gas (GS1), heater gas (GS2), and collision-activated dissociation (CAD) were set at 5500 V, 550 ◦ C, 35 psi, 40 psi, 40 psi and 7 unit, respectively. Compound-dependent parameters including declustering potential (DP), collision energy (CE), entrance potential (EP) and collision exit cell potential (CXP) were set at 50 V, 19 V, 8 V and 9 V for BBA and 41 V, 44 V, 14 V and 15 V for telmisartan. AB Sciex Analyst software (version 1.6.2) was used as system control. Watson LIMS software (version 7.5) was used for data analysis. 2.3. Preparation of working solutions, calibration standards and quality control samples 2.3.1. Method A The stock solutions of BBA (1 mg/mL), IS (1 mg/mL) and benzonatate (3 mg/mL) were all prepared in methanol. The working solutions of BBA, IS and benzonatate were prepared by consecutively diluting the stock solution with methanol/water (1:1, v/v). For BBA, the standard working solutions covered the concentration range of 200–200000 ng/mL. The concentrations of the working solutions of quality control (QC) samples were 200, 500, 5000, 150,000 and 400,000 ng/mL. The IS working solution was at the concentration of 900 ng/mL. The stock and working solutions of BBA and IS were all stored at −20 ◦ C prior to analysis. For benzonatate, the working solutions of QC samples at the concentrations of 624, 1560, 15,600, 468,000 and 1,248,000 ng/mL were prepared for the degradation study. The stock and working solutions of benzonatate were all freshly prepared when used. Calibration standards and QC samples of BBA were prepared by 20-fold spiking the working solutions into the blank plasma at room temperature. The calibration standards covered the concentration levels of 10, 20, 100, 500, 2500, 5000, 8000 and 10,000 ng/mL. The lower limit of quantification QC (LLOQ QC), low QC (LQC), medium QC (MQC), high QC (HQC) and dilution QC (DQC) were at the concentrations of 10, 25, 250, 7500 and 20,000 ng/mL, respectively. The DQC was prepared to validate the reliability of dilution process.

2.2. Instrument and HPLC–MS/MS conditions The determination of analyte was achieved on high performance liquid chromatograph (Shimadzu, Kyoto, Japan) equipped with an

2.3.2. Method B The stock solutions of BBA, IS and benzonatate were all prepared in methanol at the concentrations of 1 mg/mL, 1 mg/mL

136

J. Man et al. / Journal of Pharmaceutical and Biomedical Analysis 173 (2019) 134–143

and 3 mg/mL, respectively. Neostigmine metilsulfate accurately weighted was dissolved in water to prepare the stock solution (10 mg/mL). For both BBA and benzonatate, the preparation process and the concentrations of working solutions were consistent with that in method A. The IS working solution at the concentration of 60 ng/mL was prepared by diluting the stock solution with methanol/water (1:1, v/v). The mixed working solutions of the QC samples containing BBA and benzonatate were also prepared for the investigation of the stability. The concentrations of the mixed working solutions were 200, 500, 5000, 150,000 and 400,000 ng/mL for BBA and 624, 1560, 15600, 468,000 and 1,248,000 ng/mL for benzonatate, respectively. Calibration standards and QC samples of BBA were obtained by 20-fold spiking the working solutions into the blank plasma under ice-bath condition. And the blank plasma was centrifuged from the blood containing neostigmine metilsulfate stock solution (1000:25, v/v). The concentrations of the spiked samples were identical with that in method A. 2.4. Sample preparation 2.4.1. Sample preparation for method A Frozen samples were allowed to thaw and equilibrate to room temperature. After that, melting samples should be fully vortexed prior to analysis. Calibration standards, QC samples and drugcontaining samples were pretreated simultaneously as described below. An aliquot of 50 ␮L plasma sample was spiked with 30 ␮L of the IS working solution (900 ng/mL) and vortexed for 1 min. Then the mixture was precipitated with 370 ␮L methanol. The entire sample was thoroughly vortexed for 10 min, followed by centrifuging (4000 rpm, 10 min, 10 ◦ C). Subsequently, 30 ␮L of the supernatant was further diluted with 570 ␮L methanol/water (4:6, v/v). An aliquot of 10 ␮L of the well-mixture was injected into the HPLC–MS/MS system for analysis. 2.4.2. Sample preparation for method B Frozen samples were allowed to equilibrate to room temperature and fully vortexed before analysis. Then, all the samples were pretreated simultaneously as described below. An aliquot of 100 ␮L plasma and 800 ␮L methanol were added into a polypropylene pipe under ice-bath condition, vortexed for 10 min, then centrifuged at 12,000 rpm for 10 min at 10 ◦ C. Afterwards, an aliquot of 30 ␮L of the IS working solution (60 ng/mL) and an aliquot of 30 ␮L supernatant were successively added to 540 ␮L methanol/water (4:6, v/v) in a 96 well plate and completely vortexed for 10 min. Thereafter, 10 ␮L of the mixture was injected to the HPLC–MS/MS system for analysis. 2.5. Method development

C(Benzonatate) × V MW(Benzonatate,

average)

=



C(Benzonatate) = C(total × MW(Benzonatate,

BBA)

C(BBA

converted from

benzonatate)

MW(BBA) − C(the

×V

(2)

 original

average) /MW(BBA)

BBA)

(3)

In Eqs. (2) and (3), “C” means the concentration; “V” means the volume; “MW” means the molecular weight; MW(Benzonatate, average) = 603.74; MW(BBA) = 193.24. 2.5.1. Degradation study of benzonatate Firstly, the blood samples of benzonatate and BBA at the concentrations of LLOQ and DQC were all prepared by 20-fold spiking the working solutions into the blank blood anticoagulated by EDTA-K2 under ice-bath condition. The spiked blood samples of benzonatate were placed at room temperature for 0.5 h, 1 h, 2 h and 3 h after preparation respectively, and then centrifuged at 4000 rpm, 10 ◦ C for 10 min. As a control group, the LLOQ and DQC blood samples of BBA were centrifuged immediately. Subsequently, all the plasma samples were pretreated according to method A at once. The degradation percent of benzonatate could be evaluated by the appearance percent of BBA. And it was calculated by dividing the peak area ratios (analyte/IS) of the test samples at different time points by the average peak area ratios (analyte/IS) of the control samples. Because the concentrations of BBA in the control group were consistent with the theoretical concentrations of BBA fully converted from benzonatate in the test group. 2.5.2. Study on the blood stability conditions of benzonatate 2.5.2.1. Temperature. The blood sample was prepared by 20-fold spiking the working solution of benzonatate at DQC concentration into the blank blood anticoagulated by heparin sodium under ice-bath condition. The spiked blood sample was divided into two groups. One was placed for 0.5 h at room temperature, the other was placed for 0.5 h under ice-bath condition. After that, the blood samples were centrifuged (4000 rpm, 10 min, 10 ◦ C) and pretreated according to method B. The result was evaluated by comparing the average peak areas of BBA converted from benzonatate under the two conditions. 2.5.2.2. Anticoagulants. The working solution of benzonatate at DQC concentration was added into the blank blood anticoagulated by heparin sodium and the blank blood anticoagulated by EDTAK2 under ice-bath condition, respectively. The two kinds of blood samples were placed for 1 h under ice-bath condition and then centrifuged (4000 rpm, 10 min, 10 ◦ C). Subsequently, the plasma samples were pretreated according to method B. And the result was evaluated by comparing the average peak areas of BBA converted from benzonatate in the two kinds of blood samples.

Benzonatate was a mixture containing 15 components with different degrees of polymerization, which can hardly be directly detected by mass spectrometer. It has been confirmed that each mL of 0.5 N sodium hydroxide was equivalent to 301.5 mg (0.5 N) benzonatate in US pharmacopoeia (USP35–NF30), which indicated that each mL of 0.5 N benzonatate was equivalent to 0.5 N BBA (Eq. (1)). The concentration of benzonatate can be estimated based on the concentration of BBA converted from benzonatate (Eq. (2)). And the concentration of BBA converted from benzonatate was equal to the concentration of total BBA minus the original BBA. Therefore, the concentration of benzonatate in plasma can be calculated by the following formula (Eq. (3)). Hence, two sensitive HPLC–MS/MS methods for the determination of total BBA and the original BBA in plasma were separately developed.

2.5.2.3. Enzyme inhibitors. The neostigmine metilsulfate water solution at the concentration of 4 mg/mL and the supersaturated solution of sodium fluoride [16–19] were added into the blank blood (EDTA-K2 ), respectively. The blood samples of benzonatate at DQC concentration were prepared by 20-fold spiking the working solution into the two kinds of blank blood under ice-bath condition, separately. The two kinds of blood samples were centrifuged (4000 rpm, 10 min, 10 ◦ C) after the placement for 2 h under icebath condition and then pretreated according to method B. The inhibition effect of the two enzyme inhibitors was evaluated by comparing the average peak areas of BBA converted from benzonatate.

0.5 N(Benzonatate) = 0.5 N(BBA)

2.5.2.4. The inhibit efficiency of neostigmine metilsulfate. Firstly, 100 ␮L neostigmine metilsulfate water solution (4 mg/mL,

(1)

J. Man et al. / Journal of Pharmaceutical and Biomedical Analysis 173 (2019) 134–143

8 mg/mL, 10 mg/mL) was added into 4 mL blank blood (EDTAK2 ), separately. Subsequently, the blood samples of benzonatate were prepared by 20-fold spiking the working solutions at LLOQ and DQC concentrations into the obtained blank blood under ice-bath condition. The blood samples of BBA at corresponding concentration levels were also prepared as a control group. Finally, all the spiked blood samples were centrifuged (4000 rpm, 10 min, 10 ◦ C) after being placed for 2 h under ice-bath condition, and then pretreated according to method B. The inhibit efficiency can be evaluated by calculating the appearance percent of BBA.

137

2.6.2.2. Stability. This study was to investigate the stability of BBA in the presence of benzonatate. Hence, the blood stability samples and the plasma stability samples (LLOQ and DQC) were all prepared by 20-fold spiking the mixed working solutions into the blank matrix under ice-bath condition. The supernatant stability samples were acquired by centrifuging the plasma samples (LQC and HQC) precipitated by methanol. And the plasma samples were obtained by 20-fold spiking the mixed working solutions into the blank plasma. Room temperature stability (18.2 h), auto-sampler stability (8 ◦ C for 9 days) and freeze-thaw stability (four freezethaw cycles) in supernatant were also tested.

2.6. Method validation 2.7. Pharmacokinetic study The two methods were fully validated according to the guideline for bioanalytical method validation published by the United States Food and Drug Administration (FDA) [20]. 2.6.1. Method validation for method A 2.6.1.1. Selectivity and carry-over effect. Selectivity was evaluated by comparing the chromatograms of blank plasma samples from six different volunteers with those of blank plasma samples spiked with BBA and IS. Carry-over effect was checked by injecting a blank plasma sample directly after the injection of ULOQ samples. 2.6.1.2. Calibration curve and linearity. The calibration curves were established by plotting the peak area ratio (analyte/IS) versus the concentration of BBA. And the deviation observed in nominal and actual concentrations calculated by least-squares linear regression analysis using 1/x2 as weighing factor should be within ±15% except LLOQ (±20%). 2.6.1.3. Accuracy, precision and dilution integrity. Accuracy and precision (intra-batch and inter-batch) were determined by analyzing six replicates of QC samples at four different levels (LLOQ QC, LQC, MQC, and HQC) in three consecutive validation batches. The accuracy expressed as the relative error (RE) was required to be within ±15% (±20% for LLOQ). The precision calculated as the relative standard deviation (RSD%) should not be greater than 15% (20% for LLOQ). The dilution integrity was performed by analyzing HQC and DQC samples (n = 6) diluted by 5-fold with blank plasma. 2.6.1.4. Recovery and matrix effect. Recovery was calculated as the ratios of the peak areas of extracted samples at three concentration levels (LQC, MQC and HQC) to those of extracted blank samples spiked with neat solutions. Recovery of IS was assessed in the same way. The matrix effect was evaluated at the same concentration levels as recovery. IS-normalized matrix factors were adopted to assess the matrix effect. 2.6.1.5. Stability. The stability of BBA in whole blood and plasma was investigated by determining QC samples at two concentration levels (LQC and HQC). Stock solution stability of BBA and IS (−20 ◦ C for 15 days), auto-sampler stability (8 ◦ C for 8 days), room temperature stability (18.5 h) and freeze-thaw stability (five freeze-thaw cycles) were also considered. 2.6.2. Method validation for method B The mainly validated content of method B was consistent with that of method A. Some differences between method A and method B were briefly described as below. 2.6.2.1. Dilution integrity. The supernatant of HQC and DQC samples (n = 6) diluted by 5-fold with the supernatant of blank samples was prepared. And the supernatant was obtained by centrifuging the plasma samples precipitated by methanol.

The clinical trial was licensed by National Medical Products Administration with the registration number of 2016L06776. All the Chinese volunteers undergoing strict health check-up were recruited after signing the informed consent forms. The protocol approved by the Ethics Committee of Beijing Anzhen Hospital affiliated to Capital Medical University was obedient to the ethical principles established in the Declaration of Helsinki. A total of four Chinese volunteers including half males and half females were fasted overnight for at least 10 h before administration. Each volunteer took one benzonatate soft capsule (100 mg) orally with 240 mL warm water. All volunteers were allowed ad libitum access to water except for 1 h before and after drug administration. And they were permitted to have a standard meal until 4 h after dosing. Blood samples were collected in all the labeled vacuum tubes at 0 (pre-dosing), 0.167, 0.333, 0.5, 0.75, 1, 1.5, 2, 2.5, 3, 4, 5, 6, 7, 8 and 10 h. At each time point, 5 mL blood was collected (2 mL was drew rapidly into the precooling vacuum tube B containing 50 ␮L neostigmine metilsulfate stock solution and EDTA-K2 , the rest of the blood was drew into the vacuum tube A only containing EDTA-K2 ). Then the well-mixed blood samples in the vacuum tube B were centrifuged immediately at 4000 rpm for 10 min under low temperature condition (10 ◦ C), and an aliquot of 100 ␮L plasma was transferred into a polypropylene pipe containing 800 ␮L methanol under ice-bath condition. After that, well-vortexed samples were centrifuged at 12,000 rpm for 10 min at 10 ◦ C. The supernatant was transferred to another polypropylene pipe and stored at −70 ◦ C until analysis. The blood samples in the vacuum tube A should be placed at the room temperature for at least 2 h. Thereafter, the plasma samples were prepared by centrifuging (4000 rpm, 10 min, 10 ◦ C) and stored at −70 ◦ C prior to analysis. 3. Results and discussion 3.1. Method development 3.1.1. Results of degradation study The degradation trend of benzonatate at LLOQ and DQC concentrations was shown in Fig. 1. It was observed that the appearance percent of BBA was about 100% at 2 h. According to this conclusion, the drug-containing samples should be placed for 2 h at room temperature and then centrifuged to detect the concentration of total BBA in plasma. 3.1.2. Results of benzonatate blood stability 3.1.2.1. Temperature. The average peak area of BBA converted from benzonatate at room temperature was double that under ice-bath condition. It was indicated that low temperature can inhibit the degradation of benzonatate in blood. 3.1.2.2. Anticoagulants. After being placed for 1 h under ice-bath condition, the average peak area of BBA converted from benzonatate in the blood samples anticoagulated by heparin sodium

138

J. Man et al. / Journal of Pharmaceutical and Biomedical Analysis 173 (2019) 134–143

Fig. 1. Degradation trend of benzonatate at LLOQ and DQC concentrations (n = 3).

Fig. 3. Positive product ion mass spectra of BBA (A) and telmisartan (IS, B). Fig. 2. Inhibition profile of the degradation of benzonatate at LLOQ and DQC concentrations by neostigmine metilsulfate at different concentrations (n = 3).

was 1.2 times that in the blood samples anticoagulated by EDTA-K2 . It was showed that EDTA-K2 was better for the stability of benzonatate in blood due to the complexation of EDTA and metal ions, because metal ions can catalyze the esterase hydrolysis reaction. Although low temperature and EDTA-K2 were beneficial for the detection, the degradation of benzonatate cannot be completely inhibited. 3.1.2.3. Enzyme inhibitors. After being placed for 2 h under ice-bath condition, the average peak area of BBA converted from benzonatate in the blood samples containing sodium fluoride was 14.3 times that in the blood samples containing neostigmine metilsulfate. The results indicated that the inhibition effect of neostigmine metilsulfate was better than that of sodium fluoride. 3.1.2.4. The inhibit efficiency of neostigmine metilsulfate. After being placed for 2 h under ice-bath condition, the appearance percent of BBA in the blood samples, with 10 mg/mL neostigmine metilsulfate water solution, was the lowest. In other words, the inhibit efficiency of 10 mg/mL neostigmine metilsulfate water solution was the highest (Fig. 2). Based on the results, 10 mg/mL neostigmine metilsulfate water solution was added into the vacuum tube B in the clinic. Hence, the blank plasma was centrifuged from the blood containing neostigmine metilsulfate water solution (1000:25, v/v) and all of the

operations were under ice-bath condition except the separation of supernatant for method B. 3.1.3. Mass spectrometer BBA showed strong and steady mass spectrometric intensity in the positive mode. Obviously, the predominant protonated molecular ions [M+H]+ of BBA and IS were at m/z 194.2 and 515.3, respectively. And the product ion spectra of BBA and IS were shown in Fig. 3. Additionally, the mass spectrometer parameters including compound-dependent parameters (DP, EP, CE and CXP) and sourcedependent parameters (CUR, ionspray, GS1, GS2, TEM and CAD) were optimized to achieve higher intensity. 3.1.4. Liquid chromatography To reduce the inhibition caused by neostigmine metilsulfate on mass response, valve was used to veer strong polarity eluent to the waste. InfinityLab Poroshell 120 Phenyl Hexyl column and additive of mobile phase were key factors in improving tailing and obtaining symmetrical peak shape. Furthermore, the initial mobile phase including aqueous portion and organic portion (60:40, v/v) was optimized to prolong the retention time to avoid the matrix effect. 3.1.5. Sample preparation A simple protein precipitation method using methanol or acetonitrile was performed with sufficient sensitivity. The results showed that the recovery and the matrix effect of samples precipitated with methanol were similar to those with acetonitrile. Methanol was used for study because of better sedimentation and

J. Man et al. / Journal of Pharmaceutical and Biomedical Analysis 173 (2019) 134–143

139

Fig. 4. Typical MRM chromatograms of total BBA (I) and telmisartan (IS, II) in human plasma for method A: (A) blank plasma sample, (B) blank plasma sample spiked with BBA at 10 ng/mL and telmisartan (IS) at 900 ng/mL, and (C) a plasma sample (2250 ng/mL) obtained from a subject at 0.5 h after oral administration of benzonatate soft capsule.

lower costs. For method B, the addition of neostigmine metilsulfate stock solution in blood had no effect on the matrix effect. 3.2. Method validation 3.2.1. Method validation for method A 3.2.1.1. Selectivity and carry-over effect. The representative chromatograms of double blank sample, plasma sample spiked with BBA at LLOQ and IS and plasma sample obtained from a subject after oral administration were shown in Fig. 4. Neither co-eluting peaks nor carry-over peaks were observed at the retention time of BBA and IS. 3.2.1.2. Calibration curve and linearity. The calibration curve exhibited good linearity over the concentration range from 10 to 10,000 ng/mL with a correlation coefficient of 0.9975 or better. The typical regression equation of the calibration curve was y = 0.0004515 × C + 0.0002704. The signal-to-noise ratios (S/N) of LLOQ which can be accurately quantified were much higher than 10.

3.2.1.3. Accuracy, precision and dilution integrity. The accuracy, intra-batch precision and inter-batch precision were summarized in Table 1. It was shown that the method was reproducible and feasible for the determination of BBA in human plasma. The accuracy of dilution QC samples ranged from −6.5% to 4.1%, and RSD values were below 3.7%. 3.2.1.4. Recovery and matrix effect. The mean recovery of BBA at three QC levels and the IS at the concentration of 900 ng/mL were (88.4 ± 2.7)% and (88.6 ± 3.2)%, respectively. The IS normalized matrix factors ranged from 95.6% to 100.4% with RSD below 2.4%, indicating that there was no significant impact on the determination of BBA in different matrix. And the results were summarized in Table 2. 3.2.1.5. Stability. The stability of BBA in blood was determined for 3 h. The data of the stability of BBA in plasma was summarized in Table 3. It was indicated that BBA was stable under routine processing and storage conditions.

140

J. Man et al. / Journal of Pharmaceutical and Biomedical Analysis 173 (2019) 134–143

Table 1 Accuracy and precision data for the analysis of BBA in human plasma. inter-batch (n = 6 × 3)

intra-batch (n = 6)

Spiked conc. (ng/mL)

10 25 250 7500

Measured conc. (mean ± SD, ng/mL)

Accuracy (RE, %)

Precision (RSD, %)

Measured conc. (mean ± SD, ng/mL)

Accuracy (RE, %)

Precision (RSD, %)

10.1 ± 0.412 25.4 ± 0.969 249 ± 3.92 7510 ± 153

1.0 1.6 −0.4 0.1

4.1 3.8 1.6 2.0

9.80 ± 0.588 25.1 ± 1.05 245 ± 7.31 7630 ± 193

−2.0 0.4 −2.0 1.7

6.0 4.2 3.0 2.5

RSD–relative standard deviation; RE–relative error; n–number of replicates.

Table 2 Recovery and IS-normalized matrix factors in human plasma. Recovery

IS-normalized matrix factors

BBA (n = 6)

25 ng/mL 250 ng/mL 7500 ng/mL

BBA (n = 18)

mean ± SD (%)

RSD (%)

87.2 ± 2.2 86.5 ± 1.9 91.4 ± 1.9

2.5 2.2 2.1

mean ± SD (%) 88.4 ± 2.7

IS (n = 18) RSD (%)

3.0

mean ± SD (%) 88.6 ± 3.2

BBA (n = 6)

BBA (n = 18)

RSD (%)

mean ± SD (%)

RSD (%)

mean ± SD (%)

RSD (%)

3.6

100.4 ± 1.1 98.3 ± 1.1 95.6 ± 0.8

1.1 1.1 0.8

98.1 ± 2.4

2.4

n–number of replicates.

Table 3 Stability of BBA in plasma under various storage conditions (n = 3).

than 10, which indicated that the quantification was accurate and reliable. RE (%)

RSD (%)

23.9 ± 0.5 238 ± 2 8037 ± 70

−4.4 −4.8 7.2

2.1 0.8 0.9

25 250 7500

25.0 ± 0.6 257 ± 13 7747 ± 464

0.0 2.8 3.3

2.4 5.1 6.0

Autosampler for 8 days (8 ◦ C)

25 250 7500

24.8 ± 0.6 251 ± 19 7943 ± 425

−0.8 0.4 5.9

2.4 7.6 5.4

Freezed for 27 days (−70 ◦ C)

25 250 7500

27.6 ± 0.4 248 ± 9 7940 ± 175

10.4 −0.8 5.9

1.4 3.6 2.2

Storage conditions

Concentration levels (ng/mL) Nominal

Measured (mean ± SD)

Room temperature for 18.5 h

25 250 7500

Five freeze-thaw cycles (−70 ◦ C)

RSD–relative standard deviation; RE–relative error; n–number of replicates.

3.2.2. Method validation for method B 3.2.2.1. Selectivity and carry-over effect. The representative chromatograms of double blank sample, plasma sample spiked with BBA at LLOQ and IS and plasma sample obtained from a subject after oral administration were shown in Fig. 5. No obvious interferences and carry-over effects were observed. 3.2.2.2. Calibration curve and linearity. The method showed good linearity with a correlation coefficient of 0.9969 or better. The typical regression equation of the calibration curve calculated by least-squares linear regression analysis was y = 0.0006585 × C + 0.0001906. The signal-to-noise ratios (S/N) of LLOQ were no less

3.2.2.3. Accuracy, precision and dilution integrity. The results summarized in Table 4 suggested that the established method had a satisfactory reproducibility with RSD values less than 8.4%. The accuracy and precision of dilution QC samples were within the acceptance limits. 3.2.2.4. Recovery and matrix effect. The recovery of BBA and the IS at the concentration of 60.0 ng/mL were (99.8 ± 1.1)% and (99.6 ± 4.5)%, respectively. The IS normalized matrix factors were within the acceptance limits, indicating that determination of BBA in different matrix was reliable. And the results were summarized in Table 5. 3.2.2.5. Stability. BBA in the blood and the plasma was stable for 2 h under ice-bath condition. The data of the stability in supernatant summarized in Table 6, manifested that BBA in the presence of benzonatate was stable under various storage conditions. 3.3. Pharmacokinetic study and incurred sample reanalysis The fully validated methods were successfully applied to the pharmacokinetic study of BBA and benzonatate. Following single oral administration of benzonatate soft capsule to human volunteers, BBA and benzonatate were rapidly absorbed which reached a maximum concentration within 0.43–0.69 h post-dose. The maximum concentrations were 1708 ± 457 ng/mL for BBA and 1063 ± 460 ng/mL for benzonatate. The area under the curve from 0 h to 10 h (AUC0-10 ) for BBA and benzonatate

Table 4 Accuracy and precision data for the analysis of BBA in human plasma centrifuged from the blood containing neostigmine metilsulfate stock solution (1000:25, v/v). Spiked conc. (ng/mL)

10 25 250 7500

inter-batch (n = 6 × 3)

intra-batch (n = 6) Measured conc. (mean ± SD, ng/mL)

Accuracy (RE, %)

Precision (RSD, %)

Measured conc. (mean ± SD, ng/mL)

Accuracy (RE, %)

Precision (RSD, %)

10.2 ± 0.859 25.4 ± 0.556 243 ± 3.01 7240 ± 176

2.0 1.6 −2.8 −3.5

8.4 2.2 1.2 2.4

10.1 ± 0.656 25.0 ± 1.59 246 ± 16.1 7190 ± 557

1.0 0.0 −1.6 −4.1

6.5 6.4 6.5 7.7

RSD–relative standard deviation; RE–relative error; n–number of replicates.

J. Man et al. / Journal of Pharmaceutical and Biomedical Analysis 173 (2019) 134–143

141

Fig. 5. Typical MRM chromatograms of the original BBA (I) and telmisartan (IS, II) in human plasma centrifuged from the blood containing neostigmine metilsulfate solution for method B: (A) blank plasma sample, (B) blank plasma sample spiked with BBA at 10 ng/mL and telmisartan (IS) at 60 ng/mL, and (C) a plasma sample (1760 ng/mL) obtained from a subject at 0.5 h after oral administration of benzonatate soft capsule.

Table 5 Recovery and IS-normalized matrix factors in human plasma centrifuged from the blood containing neostigmine metilsulfate stock solution (1000:25, v/v). Recovery

IS-normalized matrix factors

BBA (n = 6)

25 ng/mL 250 ng/mL 7500 ng/mL

BBA (n = 18)

mean ± SD (%)

RSD (%)

99.8 ± 9.9 100.8 ± 6.8 98.7 ± 8.2

9.9 6.7 8.3

mean ± SD (%) 99.8 ± 1.1

IS (n = 18) RSD (%)

1.1

mean ± SD (%) 99.6 ± 4.5

BBA (n = 6)

BBA (n = 18)

RSD (%)

mean ± SD (%)

RSD (%)

mean ± SD (%)

RSD (%)

4.6

109.0 ± 1.2 108.1 ± 1.2 98.5 ± 0.6

1.1 1.1 0.6

105.2 ± 5.8

5.5

n–number of replicates.

were 2103 ± 918 ng/mL·h and 1097 ± 559 ng/mL·h, respectively. The AUC0-10 of BBA was greater than that of benzonatate, which indicated that the exposure of BBA in vivo was more than that of benzonatate. Mean plasma concentration-time profiles of BBA and benzonatate were shown in Fig. 6. The pharmacokinetic parameters of BBA and benzonatate were summarized in Table 7.

Incurred sample reanalysis was also performed. A total of 40 incurred samples (including half for method A and half for method B) near the Cmax , absorption phase and the elimination phase in the pharmacokinetic profile were reanalyzed. And the concentration differences between the original and repeated values were between −3.2–10.5 for method A and between −4.8–11.4 for method B. It

142

J. Man et al. / Journal of Pharmaceutical and Biomedical Analysis 173 (2019) 134–143

Table 6 Stability of BBA in supernatant in the presence of benzonatate under various storage conditions (n = 3). RE (%)

RSD (%)

28.4 ± 1.0 270 ± 9 7580 ± 165

13.6 8.0 1.1

3.5 3.3 2.2

25 250 7500

26.8 ± 0.4 261 ± 14 6627 ± 167

7.2 4.4 −11.6

1.5 5.4 2.5

Autosampler for 9 days (8 ◦ C)

25 250 7500

26.7 ± 1.0 255 ± 12 7480 ± 246

6.8 2.0 −0.3

3.7 4.7 3.3

Freezed for 27 days (−70 ◦ C)

25 250 7500

26.9 ± 0.9 243 ± 6 7563 ± 195

7.6 −2.8 0.8

3.3 2.5 2.6

Storage conditions

Concentration levels (ng/mL) Nominal

Measured (mean ± SD)

Room temperature for 18.2 h

25 250 7500

Four freeze-thaw cycles (−70 ◦ C)

RSD–relative standard deviation; RE–relative error; n–number of replicates.

Fig. 6. Mean plasma concentration-time profiles of BBA and benzonatate.

Table 7 Pharmacokinetic parameters of BBA and benzonatate in healthy Chinese volunteers after oral administration of benzonatate soft capsule at a single dose of 100 mg. PK parameters

BBA

Benzonatate

Tmax (h) Cmax (ng/mL) t1/2 (h) Kel (h−1 ) AUC0-10 (ng/mL·h) AUC0-∞ (ng/mL·h) CL/F (L/h) Vd/F (L)

0.56 ± 0.13 1708 ± 457 1.32 ± 0.29 0.551 ± 0.154 2103 ± 918 2140 ± 936 52.3 ± 17.5 96.0 ± 28.0

0.56 ± 0.13 1063 ± 460 1.01 ± 0.41 0.825 ± 0.466 1097 ± 559 1111 ± 576 107 ± 46 144 ± 74

was indicated that the concentrations detected by both methods were accurate and reliable. 4. Conclusion The stability of benzonatate for a certain period was solved, which made the detection of the original BBA feasible and reliable. In this study, two sensitive HPLC–MS/MS methods for the determination of total BBA and the original BBA in plasma were developed and fully validated. They were successfully applied to the pharmacokinetic study of BBA and benzonatate in human plasma after oral administration of benzonatate soft capsule. The pharmacokinetic data could provide clinical reference and fill the gap of benzonatate in pharmacokinetic profiles. Here, we first published that the concentration of trace amount of mixture of polymers in vivo can be achieved by the detection of their metabo-

lites, which provided some useful information for the similar research. Conflicts of interest The authors have declared no conflicts of interest. Acknowledgements The authors gratefully acknowledge the financial support from National Natural Science Foundation of China (No. 81573387) and (No. 81603072). This work was also supported by “Double First–Class” University project. The authors thank Yixiao Mei and Margaret Phiri for their kind help in revising the English language of this article. References [1] J. Homsi, D. Walsh, K.A. Nelson, Important drugs for cough in advanced cancer, Support. Care Cancer 9 (8) (2001) 565–574. [2] B. Estfan, S. Legrand, Management of cough in advanced cancer, J. Support. Oncol. 2 (6) (2004) 523–527. [3] B. Estfan, D. Walsh, The cough from hell: diazepam for intractable cough in a patient with renal cell carcinoma, J. Pain Symptom Manage. 36 (5) (2008) 553–558. [4] Z. Shi, T. Yu, R. Sun, S. Wang, X.Q. Chen, L.J. Cheng, R. Liu, Discovery of novel human epidermal growth factor receptor-2 inhibitors by structure-based virtual screening, Pharmacogn. Mag. 12 (46) (2016) 139–144. [5] M.L. Winter, H.A. Spiller, J.R. Griffith, Benzonatate ingestion reported to the national poison center database system (NPDS), J. Med. Toxicol. 6 (4) (2010) 398–402. [6] W.J. Lin, R.R.L. Chen, S.L. Suen, Determination of benzonatate and its metabolite in plasma by high performance liquid chromatography, Chin. Pharm. J. 48 (1996) 53–61. [7] J.D. Caldwell, B.S. Cooke, M.K. Greer, High performance liquid chromatography-size exclusion chromatography for rapid analysis of total polar compounds in used frying oils, J. Am. Oil Chem. Soc. 88 (11) (2011) 1669–1674. [8] S.J.H. Júnior, J.N.R. Ract, L.A. Gioielli, M. Vitolo, Conversion of triolein into mono- and diacylglycerols by immobilized lipase, Arab. J. Sci. Eng. 43 (5) (2017) 1–9. [9] D. Kmiecik, J. Kobus-Cisowska, B. Kulczynski, Thermal decomposition of partially hydrogenated rapeseed oil during repeated frying traditional and fast french fries, J. Am. Oil Chem. Soc. 95 (4) (2018) 473–483. [10] G. Márquez-Ruiz, F. Holgado, M.C. García-Martínez, M.C. Dobarganes, A direct and fast method to monitor lipid oxidation progress in model fatty acid methyl esters by high-performance size-exclusion chromatography, J. Chromatogr. A 1165 (1) (2007) 122–127. [11] A. Singh, Sitosterol as an antioxidant in frying oils, Food Chem. 137 (1–4) (2013) 62–67. [12] B. Ouyang, F. Zhou, L. Zhen, Y. Peng, J. Sun, Q. Chen, X. Jin, G. Wang, J. Zhang, Simultaneous determination of tenofovir alafenamide and its active metabolites tenofovir and tenofovir diphosphate in HBV-infected hepatocyte with a sensitive LC–MS/MS method, J. Pharm. Biomed. Anal. 146 (2017) 147–153. [13] B. Shen, S.Y. Li, Determination of total, free and saliva mycophenolic acid with a LC–MS/MS method: application to pharmacokinetic study in healthy volunteers and renal transplant patients, J. Pharm. Biomed. Anal. 50 (3) (2009) 515–521. [14] S. Viswanathan, P.R.P. Verma, M. Ganesan, J. Manivannan, A novel liquid chromatography/tandem mass spectrometry (LC–MS/MS) based bioanalytical method for quantification of ethyl esters of Eicosapentaenoic acid (EPA) and Docosahexaenoic acid (DHA) and its application in pharmacokinetic study, J. Pharm. Biomed. Anal. 141 (2017) 250–261. [15] L. Yuan, W. Jian, D. Zhang, A.F. Aubry, M.E. Arnold, Application of a stabilizer cocktail of N-ethylmaleimide and phenylmethanesulfonyl fluoride to concurrently stabilize the disulfide and ester containing compounds in a plasma LC–MS/MS assay, J. Pharm. Biomed. Anal. 88 (2014) 552–561. [16] P.R.N. Bhuket, N. Niwattisaiwong, P. Limpikirati, P. Khemawoot, P. Towiwat, B. Ongpipattanakul, P. Rojsitthisak, Simultaneous determination of curcumin diethyl disuccinate and its active metabolite curcumin in rat plasma by LC–MS/MS: application of esterase inhibitors in the stabilization of an ester-containing prodrug, J. Chromatogr. B 1033 (2016) 301–310. [17] M. Koitka, J. Höchel, H. Gieschen, H.H. Borchert, Improving the ex vivo stability of drug ester compounds in rat and dog serum: inhibition of the specific esterases and implications on their identity, J. Pharm. Biomed. Anal. 51 (3) (2010) 664–678. [18] X.L. Wei, R. Han, L.H. Quan, C.Y. Liu, Y.H. Liao, Stabilization of zeylenone in plasma by the presence of esterase inhibitors and its LC–MS/MS assay for pharmacokinetic study, Biomed. Chromatogr. 27 (5) (2013) 636–640.

J. Man et al. / Journal of Pharmaceutical and Biomedical Analysis 173 (2019) 134–143 [19] T. Zhou, T. Zhao, Q. Cheng, S. Liu, L. Xu, W. Tan, A sensitive LC–MS/MS method for simultaneous determination of R-bambuterol and its active metabolite R-terbutaline in human plasma and urine with application to a clinical pharmacokinetic study, Biomed. Chromatogr. 28 (7) (2014) 994–1002.

143

[20] Food and Drug Administration (FDA), Guidance for Industry-Bioanalytical Method Validation, 2018 https://www.fda.gov/downloads/Drugs/ GuidanceCompliance%20Regulatory%20Information/Guidances/UCM070107. pdf.