A high-performance liquid chromatography-tandem mass spectrometry method for the determination of lifrafenib, a novel RAF kinase and EGFR inhibitor, in human plasma and urine and its application in clinical pharmacokinetic study
Xueting Yaoa,b, Ling Songa,b, Yang Liua,b, Huanhuan Wanga,b, Jie Liua,b, Ji Jianga,b,
Dongyang Liua,b,∗, Pei Hua,b,∗∗
a Clinical Pharmacology Research Center, Peking Union Medical College Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences,
Beijing 100032, China
b Beijing Key Laboratory of Clinical PK and PD Investigation for Innovative Drugs, Beijing 100032, China
a r t i c l e i n f o a b s t r a c t
Article history:
Received 4 November 2018
Accepted 23 December 2018
Available online 24 December 2018
Keywords:
HPLC-MS/MS
Pharmacokinetics BRAF V600E EGFR
Lifirafenib
Lifirafenib (BGB-283), a dual inhibitor trageting BRAF kinase and EGFR, showed favorable efficacy and safety in treating patients with different cancer types harboring mutations in BRAF, KRAS and NRAS. In order to support the clinical pharmacokinetic study, a sensitive high performance liquid chromatography- tandem mass spectrometry (HPLC–MS/MS) method was developed and validated to quantify lifirafenib concentration in human plasma and urine. Plasma samples were purified using protein precipitation. Urine samples were pre-treated by adding tween 80 with the purpose of preventing non-specific adsorp- tion, then extracted by centrifugation. Chromatographic separation was achieved on Phenomenex Luna C18 column with a gradient elution. The mass detection was performed using electrospray ionization (ESI) source under multiple reaction monitoring (MRM) in positive ionization mode. The method was fully validated, and the result of inter-assay and intra-assay precisions were less than 15% and the accu- racy within the scope of ±15%. The linear range for plasma and urine covered from 10 to 10,000 ng/mL and 1 to 200 ng/mL, respectively, with correlation coefficients of 0.99. The validation for matrix effect, recovery, stability and carryover were met the acceptance criteria. The method showed robust and sen- sitive, it successfully fulfilled the requirement of clinical pharmacokinetic study of lifirafenib in Chinese patients with locally advanced or metastatic solid tumors.
© 2018 Published by Elsevier B.V.
1. Introduction
The rapidly accelerated fibrosarcoma (RAF) serine-threonine protein kinase family plays a crucial role in the mitogen acti- vated protein kinase (MAPK) signaling transduction pathway by mediating signals from RAS to the downstream protein kinase MEK and then activating the extracellular signal-regulated kinase (ERK1/2) [1]. Among the three members (ARAF, BRAF, and CRAF)
∗ Corresponding author at: Clinical Pharmacology Research Center, Peking Union Medical College Hospital and Chinese Academy of Medical Sciences, Beijing 100032, China. Tel.: +86 10 6915 8356; Fax: +86 10 6915 8365.
∗ ∗ Corresponding author at: Clinical Pharmacology Research Center, Peking Union
Medical College Hospital and Chinese Academy of Medical Sciences, Beijing 100032, China. Tel.: +86 10 6915 8366; Fax: +86 10 6915 8365.
E-mail addresses: [email protected] (D. Liu), hubei01 [email protected] (P. Hu).
in RAF family, BRAF has the highest kinase activity and the most frequent oncogenic mutations leading to aberrant activation of the MAPK pathway. BRAF gene mutations were detected in many human solid tumors, including approximately 50% of human malig- nant melanomas, 27% of thyroid cancers, 15% of ovarian cancers, 10%–15% of colorectal cancers, and 2%–4% non-small cell lung car- cinomas (NSCLC) [2,3]. The most prevalent BRAF mutation (more than 90%) is of valine 600 to glutamic acid (V600E), which con- fers tenfold kinase activity than wild-type isoform [4]. BRAF V600E mutation introduces a negative charge in the BRAF kinase activation loop region that mimics the phosphorylation to active BRAF kinase, thus giving rise to a constitutive MAPK signaling that promotes tumor progression [1]. BRAF V600E is an effective therapeutic target in cancer. BRAF inhibitors vemurafenib and dabrafenib have been approved for the treatment of metastatic melanoma with BRAF V600E-positive mutation. However, most patients who treated with specific BRAF inhibitor over an average of 6 months failed to
https://doi.org/10.1016/j.jpba.2018.12.038 0731-7085/© 2018 Published by Elsevier B.V.
X. Yao et al. / Journal of Pharmaceutical and Biomedical Analysis 166 (2019) 20–29 21
Fig. 1. Product ion spectrum of lifirafenib (A) and BGB-1006 (B).
have significant long-term clinical benefits largely due to the resis- tance [5,6], and postulated resistance mechanisms involved CRAF dimerization, upregulation of NRAS or high expression of epidermal growth factor receptor (EGFR) mutation [7–9].
Lifirafenib (BGB-283), the chemical formula 5-(((1R,1aS,6bR)-1-(6-(trifluoromethyl)-1H-benzo[d]imidazol-
2-yl)-1a,6b-dihydro-1H-cyclopropa[b]benzofuran-5-yl)oxy)-
3,4-dihydro-1,8-naphthyrdin-2(1 H)-one [Fig. 1 (A)], is a novel multi-target small molecular kinase inhibitor that reversibly interacts with the ATP-binding site of BRAF kinase. Lifirafenib reversibly inhibits BRAF V600E (IC50 of 23 nM), ARAF (IC50 of
5.6 nM), wild-type BRAF (IC50 of 32 nM) and CRAF (IC50 of 7.0 nM) in a time-dependent manner and also selectively inhibits EGFR kinase with IC50 ranging from 13 nM to 37 nM [10]. As a dual BRAF kinase and EGFR inhibitor, lifirafenib not only potently inhibits
BRAF V600E-driven ERK phosphorylation and cell proliferation of BRAF V600E-mutant cancer cell lines, but also inhibits the growth of tumor cells harboring EGFR mutation by blocking EGFR signaling. That combined target on BRAF and EGFR leading to greater tumor cell suppression raises the prospect of lifirafenib as a single agent in BRAF V600E-mutated cancer [10,11].
The current phase 1 clinical trial of lifirafenib was conducted in Australian patients with solid tumors. As of 31 October 2015, 31 patients (mutations: 18 KRAS; three NRAS; seven BRAF V600E; two BRAF non-V600, and one NRAS/BRAF non-V600) enrolled in seven cohorts received lifirafenib doses ranging from 5 mg QD to 60 mg QD. Among 29 evaluable patients, one of two melanoma patients with a BRAF V600E mutation had a partial response (PR). In addition, there were two confirmed PRs, including one patient with KRAS- mutated endometrial cancer and one thyroid cancer patient with
22 X. Yao et al. / Journal of Pharmaceutical and Biomedical Analysis 166 (2019) 20–29
a BRAF V600E mutation; one unconfirmed PR was observed in one patient with KRAS-mutated NSCLC [12]. As of 19 September 2016, 96 patients were enrolled in the phase 1b study of BGB-283-AU and dosed with 30 mg QD. In cohorts with previously-untreated BRAF V600E-mutated melanoma (n=7), papillary thyroid cancer (PTC) with BRAF mutations (n=3), and NSCLC with KRAS mutations (n=6), the response rates based on response evaluation criteria in solid tumors (RECIST, version 1.1) were 42.9% (95% confidence inter-
val [CI], 9.9, 81.6), 33.3% (95% CI, 0.8, 90.6) and 16.7% (95% CI, 0.4,
64.1), respectively. Lifirafenib was generally well tolerated during phase 1b study, and the most common treatment-related grade 3 treatment-emergent adverse events in phase 1b were similar to those observed in phase 1a: thrombocytopenia (14.3%), hyper- tension (11.4%), and fatigue (11.4%). Antitumor activity was not only observed in subjects with BRAF V600-mutated solid tumors including melanoma, PTC, and ovarian cancer, but also in subjects with KRAS-mutated NSCLC. Based on the favorable efficacy and safety profiles exhibited in preclinical studies and clinical study in Australian subjects, a multicenter, open-label, phase I clinical trial (NCT03641586) integrated with three stages of dose escala- tion, dose expansion, and food effect study has been initiated to evaluate the tolerability, safety, pharmacokinetics, food effect, and preliminary antitumor activities of lifirafenib for Chinese patients with locally advanced solid tumors. To facilitate the clinical study and describe the pharmacokinetic profiles of lifirafenib, a reli- able and accurate quantitative method was requisite. Therefore, a robust high performance liquid chromatography coupled with tandem mass spectrometry (HPLC-MS/MS) method was developed for the determination of lifirafenib in human plasma and urine. This method was fully validated and successfully applied to the measurement of lifirafenib in human plasma and urine samples
originating from clinical study.
2. Method
2.1. Chemicals and materials
Lifirafenib (maleate, purity 100%) and internal standard (IS) BGB-1006 (purity 96.6%) were provided by BeiGene (Beijing) Co., Ltd. (Beijing, China). Methanol and acetonitrile were of chromato- graphic grades and were purchased from Fisher Scientific (New Jersey, USA). Formic acid and tween 80 were analytical grades and were obtained from Sigma-Aldrich Co. LLC. Drug-free human plasma (anticoagulant of ethylenediamine tetraacetic acid dipotas- sium salt, EDTA-K2) and urine obtained from six different healthy volunteers were supplied by Peking Union Medical College Hos- pital (Beijing, China). Deionized water was purified using Milli-Q purification system (Bedford, USA).
2.2. LC–MS/MS conditions
The HPLC-MS/MS system consisted of Shimadzu LC-20 A sys- tem and SIL-20AC auto-sampler (Shimadzu Corporation, Kyoto, Japan) coupled to an API 4000 triple quadrupole mass spectrum (AB Sciex, CA, USA). Chromatographic separation for BGB-283 and IS was achieved on a Luna C18 column (3 µm, 2 × 50 mm) from Phenomenex (Torrance, USA) at 40 ◦C column temperature. Mobile phases consisted of water containing 0.1% formic acid (A) and ace- tonitrile and methanol (1:1 v/v) containing 0.1% formic acid (B) at a flow rate of 0.4 mL/min. The gradient elution started with A: B (5:5), linearly changed to A: B (1:9) within 1.5 min and maintained A: B (1:9) for 1 min. Then the gradient returned to initial condition of A: B (5:5) to balance the system for 1.5 min. The total run time was 4 min.
The mass spectrometer was equipped with electrospray ioniza- tion (ESI) interface. In positive ionization mode, parameters were optimized by setting an ion-spray voltage of 5000 V, an ion source temperature of 650 ◦C. Auxiliary gas and nebulizer gas both were set at 60 p.s.i, collision activated dissociation gas was set at 6 unit, while the curtain gas was 10 p.s.i. Multiple reaction monitoring (MRM) transitions were m/z 479.2 293.0 for lifirafenib, m/z 493.2
307.2 for BGB-1006, the dwell time were 200 ms and 50 ms, respectively. The optimized entrance potential and collision cell exit potential for lifirafenib and IS both were 10 V, while the declus- tering potential were 80 V and 50 V, collision energy were 42 V and 37 V, respectively. Data acquisition was performed on Analyst software version 1.5 (AB Sciex, CA, USA).
2.3. Calibration standards and quality control samples
Lifirafenib was weighted and dissolved in duplicate in dimethyl- sulfoxide (DMSO) for the preparation of stock solutions (1 mg/mL). The stock solution of IS (1 mg/mL) was prepared by weighting BGB- 1006 and dissolving in the same solvent.
Blank human plasma and urine from six healthy individuals were used to prepare calibration standards and quality control (QC) samples by diluting corresponding stock solutions. The con- centrations of lifirafenib calibration standard in plasma samples were 10, 20, 50, 200, 1000, 5000, 8000, and 10,000 ng/mL. The
lower limit of quantitation (LLOQ), low (LQC), medium (MQC), and high (HQC) QC samples were prepared at concentrations of 10, 30, 300 and 7500 ng/mL. For the preparation of urine calibra- tion standards and QC samples, working solutions (40,000 ng/mL) were separately prepared by diluting the corresponding stock solu- tions with acetonitrile-water (10:40, v/v). Then serial intermediate standard working solutions were prepared at concentrations of 20, 50, 100, 200, 500, 1000, 2000, and 4000 ng/mL, and interme- diate QC working solutions at concentrations of 20, 40, 400, and 3200 ng/mL, by dilution of corresponding working solutions with acetonitrile-water (10:40, v/v). Finally, urine calibration standards and QC samples were prepared using blank urine that contained 1% tween 80 by diluting the corresponding intermediate working solutions. The concentrations were set at 1, 2.5, 5, 10, 25, 50, 100, and 200 ng/mL in urine calibration standards, and that for urine LLOQ, LOC, MQC, and HQC were 1, 2, 20, and 160 ng/mL. With regard to IS solution used in plasma sample preparation, BGB-1006 solu- tion (50 ng/mL) was prepared by diluting IS stock solution with methanol. And IS solution at concentration of 1 ng/mL was pre- pared using acetonitrile- methanol (50:50, v/v) that was used in urine sample preparation. All stock solutions, working solutions, calibration standards, and QC samples were sealed and stored at
−80 ◦C.
2.4. Extraction procedure
Plasma samples were prepared using protein precipitation method after thawing at room temperature. A volume of 400 µL IS solution (50 ng/mL BGB-1006) that containing precipitator of methanol was added to an aliquot of 50 µL plasma sample following by thorough vortex for 30 s. After storing at 4 ◦C for at least 20 min, the sample was centrifugalized at 13,000 rpm for 10 min. Then 50 µL supernatant was collected and mixed with 600 µL reconsti- tution solution (acetonitrile-water, 50:50, v/v, 0.1% formic acid), and subsequently repeating centrifugation at 13,000 rpm for 1 min. Finally, 200 µL supernatant sample was transferred into 96-well collected plate and stored at 4 ◦C waiting for injection.
The blank urine used for calibration standards and QC prepara- tion, and clinical urine samples collected from subjects were all treated with 10% tween 80-water solution at a volume ratio of 90:10, which was a method of stopping non-specific adsorption to
X. Yao et al. / Journal of Pharmaceutical and Biomedical Analysis 166 (2019) 20–29 23
the plastic container. After that, the urine sample that containing 1% tween 80 was taken 50 µL and mixed with 500 µL IS solution (1 ng/mL BGB-1006) followed by centrifugation at 13,300 rpm for 5 min. The supernatant of 220 µL was blended in 180 µL 0.2% formic acid water solution, then transferred into 96-well collected plate and injected for LC–MS/MS analysis.
2.5. Method validation
In compliance with guidance for bioanalytical method valida- tion from China National Medical Products Administration (NMPA), US Food and Drug Administration (FDA), and European Medicine Agency (EMA) [13–15], a full validation was performed includ- ing selectivity, sensitivity, linearity, accuracy, precision, recovery, matrix effect, stability and carryover.
2.5.1. Selectivity
Double blank plasma and urine (containing tween 80) sam- ples, and blank samples spiked with internal standard from six individual healthy subjects were tested to investigate the endoge- nous interference. The LLOQ samples that were prepared using blank plasma and blank urine from six individual sources were also analyzed for selectivity. Absence of interfering component is acceptable when the peak area in the double blank samples or blank samples spiked with IS were less than 20% for lifirafenib, and 5% for BGB-1006 compared to that in LLOQ samples.
2.5.2. Linearity
The linearity of calibration curves in human plasma and urine were separately assessed for each analytical batch, while two cali- bration curves were independently prepared and distributed at the beginning and the end of batch. The linear concentration-response relationship was calculated using a 1/X2 weighted least squared linear regression.
2.5.3. Precision, accuracy
The inter-assay and intra-assay precision and accuracy were val- idated by quintuplet or sextuplet QC samples (LLOQ, LQC, MQC, and HQC) in three successive analytical runs. Coefficient of varia- tion (CV%) was used to express the relative standard deviation, and relative error (RE%) was calculated to represent accuracy between measured and nominal concentration.
2.5.4. Recovery and matrix effect
Extraction recovery and matrix effect of lifirafenib in plasma and urine samples were evaluated at three concentration levels, as same as LQC, MQC, HQC, and each level with six replicates. Recov- ery was calculated by comparing the peak area between normally prepared plasma and urine samples spiked with lifirafenib and IS in advance (group A) and lifirafenib and IS spiked after the prepara- tion of blank plasma and blank urine (group B), namely, the ratio of response in group A/response in group B was calculated as recovery. Matrix effect was validated in plasma and urine from six individ- uals by comparing the peak area between group B and standard solution that lifirafenib and IS dissolved at corresponding concen- trations (group C), thus, the ratio of response in group B/response in group C was identified as matrix effect. The recovery should be con- sistent across three concentration levels, and the inter-individual variability of matrix effect should be less than 15%.
2.5.5. Stability
The stabilities of lifirafenib in plasma and urine samples were investigated under different mimicked scenarios and validated by comparing with freshly prepared samples. The short-term and
long-term stabilities were tested by placing plasma and urine sam- ples at room temperature or in −80 ◦C freezer. The freeze-thaw
stability was tested by freezing plasma and urine samples at 80 ◦C more than 24 h then thawing them at ambient temperature, this process was repeated for at least 3 cycles. The post-preparation stability was investigated by storing the processed samples in auto- sampler at 4 ◦C for 72 h. The stability of lifirafenib and BGB-1006 stock solutions was evaluated after storage at 80 ◦C condition for several months and compared with freshly weighted and prepared stock solutions.
2.5.6. Carryover
The carryover for this current method was tested by analyzing a double blank biological matrix following the highest concentration sample in calibration standards in each analytical run. In the double blank sample, peak area should be less than 20% of that in LLOQ for the analyte, and less than 5% for IS.
2.6. Pharmacokinetic study
The HPLC-MS/MS analytical method was applied to study the pharmacokinetics characteristics of lifirafenib in a multicenter, open-label, phase I clinical trial (NCT03641586). The clinical trial was conducted completely in accordance with the Guidelines for Good Clinical Practice and the Declaration of Helsinki, and was approved by the Independence Ethics Committee at Peking Union Medical College Hospital (Beijing, China) and Beijing Cancer Hospi- tal (Beijing, China). Pharmacokinetic results in the dose escalation stage are exhibited in the present paper. Chinese patients with local advanced solid tumor who signed the Informed Consent Form (ICF) were enrolled into the dose escalation stage. The subjects orally took a single dose of lifirafenib after fasting 2 h in day 1, and fol- lowing by a two-day interval of drug discontinuance. Then, starting from day 4, subjects received drug once daily until the study end. The plasma samples were obtained before administration, at 1, 2, 3, 5, 7, 9, 24, 48, 72 h post the first dose, and at pre-dose of day 11,
day 18 and day 25, as well as 1, 2, 3, 5, 7, 9, 24 h in day 25. The urine samples were collected before dosing and at 0–4, 4–8, 8–12, 12–24, 24–48, 48–72 h after the first dose, and at pre-dose and 0–4,
48˜, 8–12, 12–24 h in day 25. In addition, urine samples were treated
with 10% tween 80 prior to the storage. All the plasma and urine samples were stored at −80 ◦C condition before analysis.
3. Results and discussion
3.1. Optimization of LC–MS/MS condition
In order to specifically detect lifirafenib and internal standard BGB-1006, the mass spectrometric conditions and chromatography settings were developed and optimized. The optimization of mass spectrometric conditions was performed on API 4000 equipped with electrospray ionization interface. By injecting standard solu- tions through a syringe pump at a flow rate of 203˜0 µL/min, the precursor ions and product ions were firstly scanned, initially by systemic automation screening, and subsequently by manual optimization. Compared with negative ionization mode, each com- pound in positive ionization mode had higher response, likely due to the presentation of basic nitrogen atoms in chemical structures that made it easier to form protonated ions [M+H]+. The prod- uct ions m/z 293.0 and m/z 307.2 that were produced by the loss of trifluoromethyl-1H-benzimidazole, were selected for MRM due to the highest abundance as shown in Fig. 1. Then, the ionization parameters such as collision energy and spray voltage, gas param- eters such as curtain gas were further optimized to acquire more stable and more sensitive response.
The chromatography conditions including stationary phase of columns, column temperature, compositions of mobile phase, and flow rate were sufficiently investigated on the purpose of achieving
24 X. Yao et al. / Journal of Pharmaceutical and Biomedical Analysis 166 (2019) 20–29
an appropriate retention time and good peak shape. After com- paring several columns with different bonded phase, Phenomenex Luna C18 column comprising octadecyl silane with ligands bound to the fully porous silica provided better retention and pretty peak shape for analytes. Meanwhile, this column was well tolerated in a wide pH range (pH 1.5–10), which provided favorable durability and stability. The column temperature was set at 40 ◦C for assuring excellent column performance and steady retention time. Isopy- knic mixture of methanol and acetonitrile as organic phase showed ideal elution for lifirafenib and IS. And 0.1% formic acid was intro- duced in aqueous phase to enhance mass spectrometric response. The gradient elution with high proportion organic phase could effectively narrow the peak broadening, and with increased pro- portion of aqueous phase, it could distinctly reduce carryover. The total run time was 4.0 min, and the retention time for lifirafenib and BGB-1006 were 2.0 min and 2.3 min, which was suitable for high throughput analysis.
3.2. Optimization of sample extraction
Precipitation using organic agents was adopted in the sam- ple preparation procedure due to its superiority of simplicity and good reproducibility. Methanol was used to precipitate protein in plasma. After fully mixing methanol with plasma, placing the mix- ture at 4 ◦C for at least 20 min was required in order to achieve the maximization of precipitation efficiency. In the case of urine sample preparation, the challenge of non-specific adsorption was successfully addressed. Urine calibration standards and QC sam- ples were prepared using blank urine and purified by direct dilution method. When calibration standards and QC samples were freshly prepared and detected in HPLC-MS/MS system, the linearity of calibration curves and accuracy of QC samples were acceptable, whereas freezing urine samples (at 4 ◦C and -80 ◦C) or storing samples at room temperature for a couple of hours resulted in unreliable results. Additionally, low recovery of lifirafenib and IS was also observed in urine analysis. The non-specific adsorption to plastic tube resulted in inconsistent loss of analytes. It is common to observe the non-specific adsorption to surfaces of containers in which urine samples are collected, stored and processed, a possible explanation is that as aqueous medium, urine is lacking capable to bind or solubilize drugs in solution [16–18]. By adding surfactant such as tween 80, tween 20, sodium dodecylbenzenesulphonate (SDBS) [18–20], adsorptive drug losses can be reduced through minimizing the interaction with container surfaces. For the max- imization of anti-adsorption efficiency and minimization of MS response suppression, the optimum content of tween 80 in urine was investigated in a series of tween 80-water solutions. The result showed that 1% tween 80 in urine sample, prepared by mixing blank urine with tween 80-water (10:90, v/v) solution at a volume ratio of 90:10, was sufficient to protect analytes from adsorptive losses and without significant impact on mass spectrometric. This approach was routinely applied in the clinical trial for collecting, transporting, and storing urine samples.
3.3. Validation results
3.3.1. Selectivity
The representative chromatograms obtained from double blank biospecimen, blank biospecimen spiked IS, LLOQ, and biological samples collected from patients with advanced solid tumor are shown in Fig. 2 (plasma) and Fig. 3 (urine). Near by the reten- tion times of lifirafenib and IS in the MRM chromatograms, it was not observed endogenous interference in the double blank samples from 6 individual human plasma and urine.
3.3.2. Linearity
The LLOQ were set at 10 ng/mL for plasma and 1 ng/mL for urine with the signal-to-noise ratio higher than 10, and it was ade- quate for pharmacokinetic study. The calibration curves for human plasma and urine spiked BGB-283 were approved to be linear over the range of 10-10,000 ng/mL and 1–200 ng/mL, respectively. After regression using 1/X2 weighted least square method, excel- lent linearity was observed in each analytical run with correlation coefficient (r) more than 0.99. The back-calculated concentrations of calibration standards and linear equation parameters (slope and intercept) for validation batches are summarized in Table 1.
3.3.3. Precision and accuracy
Three consecutive analysis batches were performed to validate precision and accuracy at four concentration levels. The intra-assay and inter-assay precision and accuracy are listed in Table 2. The precision and accuracy were wet the acceptance criteria, which demonstrated the excellent repeatability and reliability for this method.
3.3.4. Recovery and matrix effect
The mean observed recoveries for the plasma extraction method were 85.3%, 86.9%, and 89.6% for LQC, MQC, and HQC levels, while that for the urine extraction method were ranged from 95.4% to 103.3%. Recoveries for lifirafenib were comparable and consistent at three concentration levels. In the present method, plasma and urine matrix obtained from six individuals showed no significant ion enhancement or suppression on the analytes, and inter-subject variability was less than 10%. The details for recovery and matrix effect are showed in Table 3.
3.3.5. Stability
The stabilities for lifirafenib in human plasma and urine as well as in stock solution are summarized in Table 4. The valida- tion results indicated that lifirafenib in human plasma and urine remained stable after being placed at ambient temperature for 18 h and 7 h, and also stable after being stored at 80 ◦C for 21 months and 15 months, respectively. Furthermore, the plasma and urine (after the addition of tween 80) samples could tolerate at least 3 freeze-thaw cycles by freezing at 80 ◦C for more than 24 h and thawing at room temperature. The post-preparation samples were placed in auto-sampler (4 ◦C) for at least 72 h, which had no impact
on the accuracy of quantification. The stock solutions of lifirafenib and IS were stably preserved at −80 ◦C for at least 14 months.
3.3.6. Carryover
In each analytical run, the chromatogram of double blank fol- lowing the upper limit of quantification was showed no peak interference around the retention time of analytes, the carryover was met the acceptable criteria.
3.4. Application to a pharmacokinetic study
In the multicenter, open-label clinical trial of lifirafenib in Chi- nese subjects with locally advanced or metastatic malignant solid tumor, this validated LC MS/MS method was successfully applied in pharmacokinetic study. After a single oral dose of 10 mg or 15 mg, plasma concentration of lifirafenib increased rapidly and reached the maximum concentration (Cmax) around a median of 2–3 h. In most subjects, second peak concentrations were observed around 71˜0 h.. The plasma concentration-time profiles of lifirafenib in 10 mg and 15 mg groups are depicted in Fig. 4. The phar- macokinetic parameters from non-compartmental analysis using Pheonix WinNonlin (Version 8.1, Certara, MO, USA) are shown in Table 5. Lifirafenib exhibited the characteristics of large inter- individual variability and low elimination. After administration
X. Yao et al. / Journal of Pharmaceutical and Biomedical Analysis 166 (2019) 20–29 25
Fig. 2. Chromatograms of lifirafenib and BGB-1006 in.(A) double blank plasma, (B) blank plasma spiked with internal standard, (C) LLOQ at 10 ng/mL, and (D) 2h plasma sample after administration 10 mg lifirafenib in a patient with advanced solid tumor (plasma concentration 205 ng/mL).
26 X. Yao et al. / Journal of Pharmaceutical and Biomedical Analysis 166 (2019) 20–29
Fig. 3. Chromatograms of lifirafenib and BGB-1006 in.(A) double blank urine, (B) blank urine spiked with internal standard, (C) LLOQ at 1 ng/mL, and (D) 0–4 h urine sample after administration 15 mg lifirafenib in a patient (urine concentration 1.48 ng/mL).
X. Yao et al. / Journal of Pharmaceutical and Biomedical Analysis 166 (2019) 20–29 27
Table 1
Back-calculated lifirafenib concentrations in plasma and urine calibration standards (Linear weighted 1/x2 ).
Nominal concentration (ng/mL) 10.0 20.0 50.0 200 1000 5000 8000 10,000 Slope Intercept R2
Mean a 9.91 20.2 51.3 193 1030 4960 8010 9850 0.002 0.002 0.995
Plasma (n = 28) CV% b
8.1 6.5 5.7 5.6 3.4 4.1 4.0 4.2 50 150 0.2
RE% c −0.9 1.0 2.6 −3.5 3.0 −0.8 0.1 −1.5
Nominal concentration 1.00 2.50 5.00 10.0 25.0 50.0 100 200 Slope Intercept R2
(ng/mL)
Mean a 0.997 2.51 5.02 10.0 25.4 51.8 98.1 191 0.069 0.007 0.997
Urine (n = 14) CV% b
3.9 5.1 4.9 4.9 2.9 5.0 3.1 3.5 5.8 85.7 0.3
RE% c −0.3 0.4 0.4 0 1.6 3.6 −1.9 −4.5
a Mean concentrations in all validation batches.
b Coefficient of variation = standard deviation/mean × 100 (in one decimal place).
c Relative error= [(mean measured concentration − nominal concentration)/(nominal concentration)] × 100 (in one decimal).
Table 2
Intra-assay and inter-assay precision and accuracy of lifirafenib in human plasma and urine.
Plasma QC levels Nominal concentration (ng/mL) Measured concentration (ng/mL) Precision (CV%) Accuracy (RE%)
Intra-assay (n = 5) LLOQ LQC MQC 10.0
30.0
300 9.49
28.7
317 9.8
7.1
3.0 −5.1
−4.3
5.7
Inter-assay (n = 15)
HQC 7500 7610 2.7 1.5
LLOQ 10.0 9.40 10.7 −6.0
LQC 30.0 28.6 6.0 −4.7
Table 3
Recovery and matrix effect for lifirafenib and IS in human plasma and urine.
Plasma QC levels Nominal concentration
Recovery% Matrix Effect%
(ng/mL) Mean CV% Mean CV%
LQC 30.0 85.3 6.5 106.8 6.4
lifirafenib (n = 6) MQC 300 86.9 2.9 105.6 1.4
HQC 7500 89.6 2.6 104.1 1.1
IS (n = 6) 50.0 79.0 3.2 97.4 2.3
Urine QC levels Nominal concentration Recovery% Matrix Effect%
(ng/mL) Mean CV% Mean CV%
LQC 2.00 103.3 5.0 100.2 5.2
lifirafenib (n = 6) MQC 20.0 100.3 2.4 107.1 3.8
HQC 160 95.4 1.8 109.8 3.1
IS (n = 6) 1.00 96.7 5.1 99.8 5.5
lifirafenib once daily for 22 consecutive days, the exposure within 24 h (AUC0-24 h) in day 25 was significantly higher than that in day 1, and the accumulation ratio was approximate 600%. In the most sub- jects, urinary concentrations of lifirafenib within 72 h after the first dose were below the limit of quantification, and the concentrations in day 25 were less than 10 ng/mL. The percentage of cumulative urine excretion of lifirafenib was less than 1%, which indicted that renal excretion might not be a main elimination route.
3.5. Incurred samples reanalysis
Many factors, including differences in instance hemolysis, protein binding, may influence the accuracy of analyte in the actual study samples. Therefore, incurred samples reanalysis (ISR) was performed by reanalyzing at least 10% of the sam- ples. More than 90% ISR samples were within the acceptance criteria that percent difference between initial concentrations and reanalysis concentrations was not greater than 20% of their
28 X. Yao et al. / Journal of Pharmaceutical and Biomedical Analysis 166 (2019) 20–29
Table 4
Stability validation for lifirafenib in human plasma and urine, and stock solutions (n = 6).
Plasma QC levels Nominal concentration (ng/mL) Measured concentration (ng/mL) CV% RE%
Short-term stability at room temperature for 18 h
Long-term stability at −80 ◦C for 21 months
Freeze-thaw stability for 3 cycles
Post-preparation stability in auto-sampler at 4 ◦C for 72 h
LQC 30.0 29.2 5.5 −2.7
MQC 300 295 2.0 −1.5
HQC 7500 7440 3.0 −7.0
LQC 30.0 31.5 11.4 5.1
MQC 300 304 2.6 1.2
HQC 7500 7127 2.3 −10.9
LQC 30.0 30.8 4.2 2.7
MQC 300 295 3.2 −1.7
HQC 7500 7024 3.9 −12.2
LQC 30.0 30.0 4.8 −0.1
MQC 300 320 3.6 6.5
BGB-1006 at −80 ◦C for 14 months
Table 5
Non-compartment pharmacokinetic parameters of lifirafenib after oral administration 10 mg and 15 mg (mean ± standard deviation).
Pharmacokinetic parameters
10 mg (N = 3) 15 mg (N = 11)
day 1 day 25 day 1 day 25
Cmax (ng/mL) 222 ± 78 1170 ± 210 373.36 ± 120.18 32950 ± 12,830
Tmax *(h) 3 (2, 5) 1 (1, 2) 2 (1. 9) 2 (0, 24)
AUC0-24h (ng/mL*h) 3547.6 ± 626.3 22494.9 ± 2242.8 5890.9 ± 1825.7 32950.0 ± 12,830.0
AUC0-t (ng/mL*h) 8003.2 ± 1467.5 22494.9 ± 2242.8 16482.1 ± 5353.8 32950.0 ± 12,830.0
Cmax: maximum plasma concentration; * Tmax : time of maximum plasma concentration, median (min, max); AUC0–24h: area under the plasma concentration-time curve from zero to 24 h; AUC0–t: area under the plasma concentration-time curve from zero to time hour (AUC0–72 for day 1 and AUC0–24 for day 25).
Fig. 4. Plasma concentration-time profiles in patients after a single oral adminis- tration of 10 mg and 15 mg lifirafenib (mean + standard deviation).
mean. It confirmed the reliability and efficiency for the present method.
4. Conclusion
This is the first description of a sensitive HPLC-MS/MS analyti- cal method developed for the quantification of lifirafenib in human plasma and urine sample. This method was strictly validated in the light of guidance requirements, and proved to be robust with good precision and accuracy. The method was successfully applied for the pharmacokinetic study in the clinical trial for lifirafenib in Chi- nese subjects with advanced solid tumors. The pharmacokinetic profiles obtained with this methodology were used to support the recommended dose to be used for further testing in this popula- tion.
X. Yao et al. / Journal of Pharmaceutical and Biomedical Analysis 166 (2019) 20–29 29
Conflicts of interest
No commercial organizations had any role in the writing of this paper for publication. The authors declare no conflicts of interest with commercial organizations.
Acknowledgments
The study was supported by the “13th Five-Year” National Major New Drug Projects of China (Ministry of Science and Technology of the People’s Republic of China, No. 2017ZX09304031-001 and No. 2017ZX09101001-002-001) and BeiGene (Beijing) Co., Ltd (Beijing, China).
References
[1] W. Fiskus, N. Mitsiades, B-Raf inhibition in the clinic: present and future, Annu. Rev. Med. 67 (2016) 29–43.
[2] H. Davies, G.R. Bignell, C. Cox, P. Stephens, S. Edkins, S. Clegg, J. Teague, H. Woffendin, M.J. Garnett, W. Bottomley, N. Davis, E. Dicks, R. Ewing, Y. Floyd, K. Gray, S. Hall, R. Hawes, J. Hughes, V. Kosmidou, A. Menzies, C. Mould, A. Parker, C. Stevens, S. Watt, S. Hooper, R. Wilson, H. Jayatilake, B.A. Gusterson,
C. Cooper, J. Shipley, D. Hargrave, K. Pritchard-Jones, N. Maitland, G. Chenevix-Trench, G.J. Riggins, D.D. Bigner, G. Palmieri, A. Cossu, A. Flanagan,
A. Nicholson, J.W. Ho, S.Y. Leung, S.T. Yuen, B.L. Weber, H.F. Seigler, T.L. Darrow, H. Paterson, R. Marais, C.J. Marshall, R. Wooster, M.R. Stratton, P.A. Futreal, Mutations of the BRAF gene in human cancer, Nature 417 (6892) (2002) 949–954.
[3] R.B. Corcoran, J. Settleman, J.A. Engelman, Potential therapeutic strategies to overcome acquired resistance to BRAF or MEK inhibitors in BRAF mutant cancers, Oncotarget 2 (4) (2011) 336–346.
[4] T. Liu, Z. Wang, P. Guo, N. Ding, Electrostatic mechanism of V600E mutation-induced B-Raf constitutive activation in colorectal cancer: molecular implications for the selectivity difference between type-I and type-II inhibitors, Eur. Biophys. J. (2018).
[5] A.X. Torres-Collado, J. Knott, A.R. Jazirehi, Reversal of resistance in targeted therapy of metastatic melanoma: lessons learned from Vemurafenib (BRAF(V600E)-specific inhibitor), Cancers 10 (6) (2018).
[6] J. Charles, C. Martel, F. de Fraipont, M.T. Leccia, C. Robert, B. Busser, Mechanisms of resistance to anti-BRAF treatments, Annales de dermatologie et de venereologie 141 (11) (2014) 671–681.
[7] C. Montagut, S.V. Sharma, T. Shioda, U. McDermott, M. Ulman, L.E. Ulkus, D. Dias-Santagata, H. Stubbs, D.Y. Lee, A. Singh, L. Drew, D.A. Haber, J. Settleman, Elevated CRAF as a potential mechanism of acquired resistance to BRAF inhibition in melanoma, Cancer Res. 68 (12) (2008) 4853–4861.
[8] R. Nazarian, H. Shi, Q. Wang, X. Kong, R.C. Koya, H. Lee, Z. Chen, M.K. Lee, N. Attar, H. Sazegar, T. Chodon, S.F. Nelson, G. McArthur, J.A. Sosman, A. Ribas,
R.S. Lo, Melanomas acquire resistance to B-RAF(V600E) inhibition by RTK or N-RAS upregulation, Nature 468 (7326) (2010) 973–977.
[9] N. Bardeesy, M. Kim, J. Xu, R.S. Kim, Q. Shen, M.W. Bosenberg, W.H. Wong, L. Chin, Role of epidermal growth factor receptor signaling in RAS-driven melanoma, Mol. Cell. Biol. 25 (10) (2005) 4176–4188.
[10] Z. Tang, X. Yuan, R. Du, S.H. Cheung, G. Zhang, J. Wei, Y. Zhao, Y. Feng, H. Peng,
Y. Zhang, Y. Du, X. Hu, W. Gong, Y. Liu, Y. Gao, Y. Liu, R. Hao, S. Li, S. Wang, J. Ji,
L. Zhang, S. Li, D. Sutton, M. Wei, C. Zhou, L. Wang, L. Luo, BGB-283, a Novel RAF Kinase and EGFR inhibitor, displays potent antitumor activity in
BRAF-mutated colorectal cancers, Mol. Cancer Ther. 14 (10) (2015)
2187–2197.
[11] H. Kotani, Y. Adachi, H. Kitai, S. Tomida, H. Bando, A.C. Faber, T. Yoshino, D.C. Voon, S. Yano, H. Ebi, Distinct dependencies on receptor tyrosine kinases in the regulation of MAPK signaling between BRAF V600E and non-V600E mutant lung cancers, Oncogene 37 (13) (2018) 1775–1787.
[12] BGB-283 deemed effective in phase I study, Cancer Discov. 6 (7) (2016) OF1.
[13] European Medicine Agency (EMA), Guideline on Bioanalytical Method Validation, Available at:, 2012 http://www.ema.europa.eu/docs/en GB/ document library/Scientific guideline/2011/08/WC500109686.pdf.
[14] Food and Drug Administration (FDA), Guidance for Industry Bioanalytical Method Validation, Center for Drug Evaluation and Research (CDER).
[15] China Food and Drug Administration (CFDA), Guideline on Bioanalytical Method Validation, China Pharmacopeia (Version 2015), Part 2, 2015, pp. 363–368.
[16] S. Silvester, F. Zang, Overcoming non-specific adsorption issues for AZD9164 in human urine samples: consideration of bioanalytical and metabolite identification procedures, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 893-894 (2012) 134–143.
[17] W. Li, S. Luo, H.T. Smith, F.L. Tse, Quantitative determination of BAF312, a S1P-R modulator, in human urine by LC-MS/MS: prevention and recovery of lost analyte due to container surface adsorption, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 878 (5-6) (2010) 583–589.
[18] Y. Xu, L. Du, M.J. Rose, I. Fu, E.J. Woolf, D.G. Musson, Concerns in the development of an assay for determination of a highly conjugated adsorption-prone compound in human urine, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 818 (2) (2005) 241–248.
[19] C. Chen, L. Bajpai, N. Mollova, K. Leung, Sensitive and cost-effective LC-MS/MS method for quantitation of CVT-6883 in human urine using sodium dodecylbenzenesulfonate additive to eliminate adsorptive losses, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 877 (10) (2009) 943–947.
[20] M.D. Anderson, S.A. Breidinger, E.J. Woolf, Effect of disease state on ionization during bioanalysis of MK-7009, a selective HCV NS3/NS4 protease inhibitor, in human plasma and human Tween-treated urine by high-performance liquid chromatography with tandem mass spectrometric detection, J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 877 (11-12) (2009) 1047–1056.