Pharmacokinetic Study and Tissue Distribution of Lorlatinib in Mouse Serum and Tissue Samples by Liquid Chromatography-Mass Spectrometry

In the present study, we developed and validated a rapid and simple liquid chromatography-tandem mass spectrometry (LC-MS/MS) method for the determination of lorlatinib in mouse serum and tissue samples, and such a method was successfully applied to investigate the pharmacokinetic study and tissue distribution of lorlatinib after oral administration. Samples were processed with methanol to precipitate protein and extract drugs, and Afatinib-d6 was used as the internal standard (IS). For LC-MS/MS analysis, compounds were separated on a C18 column by gradient elution (0.1% of formic acid and methanol) at 0.5 mL/min in the positive-ion mode with m/z 407.28 [M + H]+ for lorlatinib and m/z 492.10 [M + H]+ for IS. Good linearity was observed within the calibration ranges. Selectivity, accuracy (−6.42% to 8.84%), precision (1.69% to 10.98%), recoveries (91.4% to 115.0%), and matrix effect (84.2% to 110.6%) were all within the acceptable ranges. After oral administration, serum concentration of lorlatinib quickly achieved the maximal concentration (2,705.683 ± 539.779 μg/L) at 0.625 ± 0.231 h. The highest concentration was detected in the liver (3,153.93 ng/100 mg), followed by the stomach (2,159.92 ng/100 mg) and the kidney (548.83 ng/100 mg). In conclusion, a simple and rapid detection method was established and validated for determination of lorlatinib in blood and tissue samples of mouse. The pharmacokinetic study and tissue distribution of lorlatinib were successfully investigated using this method.


Introduction
Crizotinib, a first-generation anaplastic lymphoma kinase (ALK) inhibitor, and ceritinib and alectinib, the secondgeneration ALK inhibitors, have been used as the first-line treatment for non-small-cell lung cancer (NSCLC) carrying ALK or ROS1 rearrangements [1]. However, there are two important clinical issues in ALK-positive NSCLC treatment using ALK inhibitors. Brain metastases are commonly detected in ALK-rearranged NSCLC [2], and the cumulative incidence of BM is 58.4% at 3 years [3]. Preventing BM is the first challenge in ALK-positive NSCLC treatment. e firstgeneration ALK-tyrosine kinase inhibitors (TKIs) have poor central nervous system penetration, and the intracranial efficacy of second-generation ALK-TKIs have been improved [4].
Besides, the second challenge is that several resistance mechanisms, such as secondary mutation resistance, and activation of bypass signaling pathways may cause drug resistance during tumor treatment [5]. To control intracranial disease and mutation resistance, lorlatinib is developed as a third-generation TKI that targets ALK and ROS1 [1]. Moreover, in vitro assays exhibit that lorlatinib has greater potency and selectivity than previous generations of ALK-TKIs [6,7]. More importantly, it remains highly active in patients who have been treated with previous generations of ALK-TKIs [8]. In addition, lorlatinib can induce intracranial responses and cause neurological sideeffects due to its high brain penetration ability [8].
Few bioanalytical assays have been described for the quantification of lorlatinib in mouse serum. In previous studies, protein precipitation is used as a pretreatment step and liquid chromatography-tandem mass spectrometry (LC-MS/MS) determination is applied to quantify lorlatinib [9][10][11]. However, there is no method available for quantification of lorlatinib in tissues containing more endogenous substances. e distribution of lorlatinib in the brain is crucial for controlling brain metastases. Nevertheless, there are no in vivo studies available about the biodistribution of lorlatinib. erefore, it is crucially necessary to establish a simple and sensitive method for quantification of lorlatinib in serum and tissue samples.
In the present study, we developed a simple and sensitive LC-MS/MS method for determination of lorlatinib in mouse serum and tissue samples. Water applied for other experiments was purified by reverse osmosis. All of other reagents were of analytical grade unless otherwise indicated.

Equipment.
e LC-MS/MS system was composed of the chromatographic system, consisting of two Accela pumps (ACQUITY UPLC I-CLASS BSM), an autosampler (ACQUITY UPLC I-CLASS SM-FIN) and a column oven (ACQUITY UPLC I-CLASS CH-A), and a Xevo TQ-S mass spectrometer equipped with heated electrospray ionization (Waters, USA). e LC-MS/MS was performed using the MassLynx software (version 4.1). Briefly, 2 μL pretreated sample was injected into an Aquity UPLC ® BEH C18 column (2.1 mm × 50 mm, dp � 1.7 μm, Waters, USA), which was protected by an In-line Filter Assembly (ACQUITY GUARD FILTER, USA). H1650R desktop high-speed refrigerated centrifuge was purchased from Shanghai Lu Xiang Yi Centrifuge Instrument Co., Ltd. (China). BT125D electronic balance was provided by Sartorius Lab Instruments GmbH & Co. KG (Germany). e G560E vortex mixer was obtained from Scientific Industries (USA). e XINW-M48 high-throughput tissue homogenizer was supplied from Shanghai Xin Weng Scientific Instrument Co., Ltd. (China).

HPLC-MS Conditions.
e mobile phase consisted of 0.1% formic acid solution (A) and methanol (B), and the flow rate of the mobile phase was set at 0.5 mL/min. e column oven was maintained at 50°C. e gradient elution program was set as follows: 0-0.  Table 1. e structures and typical mass spectra of lorlatinib and IS in the positive-ion mode are illustrated in Figure 1.

Animals.
All the animal-related experiments were performed in accordance with guidelines of Institutional Experimental Animal Ethical Committee. Kun Ming mice (18-20 g) were obtained from the Beijing HFK Bioscience Co., Ltd. (Beijing, China). All mice were given free access to normal diet and water during the experiment with an exception that mice were fasted for 12 h prior to drug administration.

Preparation of Calibration Standards and Quality Control
(QC) Samples. Lorlatinib was dissolved in methanol to obtain a primary stock solution of 1.0 mg/mL and kept at 4°C prior to further analysis. Working standard solutions were prepared by diluting the primary stock solution with methanol to a final concentration of 100.0 μg/mL. Serum and tissue calibration standards were prepared by adding working standard solutions to blank mouse serum and methanol extraction of blank tissue homogenate. Tested tissues included the skeletal muscle, fat, cerebrum, heart, liver, spleen, lung, kidney, stomach, small intestine, and large intestine. Eight concentrations of serum calibration standards were 1,000, 500, 200, 100, 50, 20, 10, and 5 ng/mL. Seven concentrations of tissue calibration standards were 500, 200, 100, 50, 20, 10, and 5 ng/ mL. Serum and tissue QC samples were produced in the same way as abovementioned. Four concentrations of serum QC samples were 800 (QCH), 100 (QCM), 10 (QCL), and 5 (LLOQ) ng/mL. Four concentrations of tissue QC samples were 400 (QCH), 100 (QCM), 10 (QCL), and 5 (LLOQ) ng/ mL. Calibration standards and QC samples were stored at −20°C prior to further analysis.

Serum Sample Pretreatment.
In order to make all serum samples fall within the range suitable for testing, 10 μL of serum sample was diluted to 100 μL with blank serum. e diluted serum sample was evenly mixed using vortexmixing. A mixture sample was prepared by adding 200 μL of IS working solution (10 ng/mL) to 20 μL of diluted serum sample, and the mixture was vortex-mixed for 3 min. In order to extract lorlatinib, the supernatant was obtained by centrifugation (11,000 rpm, 15 min, 4°C). en, 100 μL of supernatant was diluted to 200 μL with purified water, and the mixture was immediately vortex-mixed for 3 min.

Tissue Sample Pretreatment.
Tissue samples were weighed, and methanol was added to tissue samples at a ratio of methanol: tissue � 2 mL: 100 mg. Subsequently, tissue samples were homogenized using a high-throughput tissue homogenizer and vortex-mixed for 3 min. Supernatant was obtained by centrifuging at 11,000 rpm for 15 min. en, 400 μL of 40 ng/mL IS solution was added to 100 μL supernatant. After vortex-mixing, supernatant was isolated by centrifugation.

Calibration.
For blood sample and each tissue sample, the x-y scatter plot consisting of eight different points was drawn in accordance with the quantification of calibration standards. In this plot, the y-axis measured the ratio of the analyte and IS and x-axis determined the lorlatinib concentration. e linear calibration curve was calculated by the regression model based on the x-y plot. For the nominal value, the total allowable accuracy and precision were within 15%. For the lower limit of quantitation (LLOQ), the allowable accuracy and precision were within 20%.

Accuracy and Precision.
Accuracy refers to how close the detection value was to true value, and it was described by relative error (RE%). Precision refers to the magnitude of random errors of measurements, and it was expressed by relative standard deviation (RSD%). e within-day accuracy and precision were expressed by analyzing the data set obtained from replicated QC samples. e between-day accuracy and precision were calculated by the data set obtained from repeated experiments (including the production, pretreatment, and quantification of the QC samples) during 3 consecutive days. e acceptable withinday and between-day accuracy and precision should not exceed 15%, while they were less than or equal to 20% for LLOQ.

Recovery and Matrix Effect.
e extraction recovery was expressed at high, medium, and low QC levels by calculating the corrected peak area ratio of blank samples spiked with lorlatinib before extraction to blank samples spiked with lorlatinib after extraction. e matrix effect (serum and tissue homogenates) was assessed by the postextraction spike method. Matrix effect was described by the corrected peak area ratio of blank samples spiked with lorlatinib after extraction to the solution of lorlatinib at equivalent concentration prepared in mobile phase.

Pharmacokinetic Study and Tissue Distribution.
e mice were orally administered with 10 mg/kg lorlatinib. Blood and tissue samples were collected from mice at 0.5, 1, 2, 4, 8, and 24 h after administration. e blood samples were transferred into glass containers and clotted at room temperature for 1 h. Once the clot was formed, blood sample was centrifuged at 4,000 rpm for 10 min, and supernatant was collected. e serum was transferred into another tube and stored at −80°C prior to further analysis. Tissue samples were rinsed by physiological saline, and then excess physiological saline was blotted up. e tissue samples were transferred into PP plastic bags and stored at −80°C until further detection.

Data Analysis.
e pharmacokinetic parameters of lorlatinib were estimated with noncompartment model by DAS (Data Analysis System) software (version 3.0, BioGuider Co., Shanghai, China). Moment analysis was used as a noncompartmental approach to obtain various pharmacokinetic parameters. Noncompartmental approach allowed the estimation of pharmacokinetic parameters without assuming any structural properties of the pharmacokinetic behavior of lorlatinib.
e area under the curve (AUC) and the first moment curve (AUMC) were calculated by using the linear trapezoidal method, which is required by FDA. e half-life (t1/2) was estimated as MRT multiplied by 0.693. C max was defined as maximal concentration, and T max was defined as the time at which the C max was observed. e tissue distribution was described by quantifying lorlatinib in various tissues at 0.5, 1, 2, 4, 8, and 24 h after administration. All data were summarized to calculate means and standard deviations, and then the concentration-time curve was plotted.

Conditions for MS and HPLC.
Conditions for MS and HPLC initially were referred to existing literature for lorlatinib [9]. In order to determine the lorlatinib concentration in tissues, chromatographic and mass spectrometric conditions were optimized according to multiple considerations, such as retention time, ionization behavior of IS, and proteins and metabolites in tissue samples. Considering the accuracy and precision of analysis in the pilot experiments, Afatinib-d6 was selected as the IS for further method validation and drug quantification. e positive-ion mode was chosen at m/z 407.28 and m/z 492.10 for lorlatinib and IS, respectively.

Selectivity.
No interfering peaks were shown in the analysis of blank blood samples (n � 6) and tissue samples (n � 6). Figure 2 illustrates representative chromatograms of blank samples, blank samples spiked with lorlatinib and IS, and samples after lorlatinib administration. Response of blank samples was compared with that of the LLOQ samples. e signal noise of lorlatinib was less than 20% of LLOQ response level of lorlatinib. e signal noise of IS was less than 1% of the LLOQ response level of IS. Accordingly, the analyte and IS would not be interfered by endogenous components while using the newly developed method for analysis.

Linearity.
For blood sample and each tissue sample, the linear calibration curve was calculated by linear regression model. Table 2 presents the linear equation of calibration curves, as well as its linear regression coefficient and linear ranges. Linearity of all calibration curves over the concentration ranges met the requirements (R 2 > 0.99) of the pharmacokinetic study. All LLOQs were below the minimum concentration required for analysis, demonstrating the allowance of lorlatinib quantitation at 24 h after administration. Table 3 shows the accuracy and precision for blood and tissue samples. All within-day accuracy and between-day accuracy at three different concentrations ranged from −6.42% to 8.84%, and precision ranged from 1.69% to 10.98%, indicating that the precision and accuracy of the HPLC-MS/MS analysis method met the acceptable criteria [12].

Recovery and Matrix Effect.
All recoveries were within the range of 91.4%-115.0%, suggesting no significant extraction losses in the pretreatment of samples. e matrix effect of lorlatinib ranged from 84.2% to 110.6%. e results of matrix effect study demonstrated that no significant response could be observed at the retention time of lorlatinib. Table 4 shows the recovery and matrix effect.

Pharmacokinetic Study and Tissue Distribution.
Noncompartmental models were used to describe pharmacokinetics of lorlatinib. Figure 3 shows the mean serum concentration versus time curve.  Figure 4 reveals that lorlatinib was rapidly and widely distributed in different tissues. In each tissue, the maximum concentration of lorlatinib appeared at 0.5 h. At this time point, the highest concentration was found in the liver (3,153.93 ng/ 100 mg), followed by the stomach (2,159.92 ng/100 mg), kidney (548.83 ng/100 mg), small intestine (540.64 ng/100 mg), heart (367.31 ng/100 mg), lung (309.39 ng/100 mg), large intestine (289.85 ng/100 mg), spleen (246.81 ng/100 mg), and brain (185.03 ng/100 mg). e first pass effect of oral administration might be the main cause of high distribution in the liver. e high concentration in the stomach and small intestines suggested that lorlatinib was primarily absorbed by the stomach and small intestines. Although lorlatinib, a third-generation ALK inhibitor, is specifically designed to penetrate the bloodbrain barrier, there was no expected high concentration of lorlatinib in the brain. In contrast, there was a higher distribution of lorlatinib in muscle tissues, such as heart, indicating that the lipophilicity of lorlatinib was not high enough to help the drug cross the blood-brain barrier. e kidneys had high    concentration of lorlatinib, implying that lorlatinib might be excreted by the kidneys. High concentration of lorlatinib in the liver could be probably attributed to a slow metabolism in the liver, leading to a high lorlatinib concentration in the liver.

Conclusions
In the present study, we established a simple and rapid detection method and validated such method for lorlatinib determination in blood and tissue samples of mice. e pharmacokinetic study and tissue distribution were successfully investigated using this method. e high concentration of lorlatinib in the heart suggested that the lipophilicity of the drug was not as high as expected, resulting in low concentration in brain, even though lorlatinib is designed to penetrate the blood-brain barrier. High concentration of lorlatinib in the kidneys and livers implied that the kidney was the primary organ to eliminate lorlatinib, and metabolism in the liver was slow. e results of the pharmacokinetic study and tissue distribution provided valuable insights into the establishment of physiologically based pharmacokinetic model and its further evaluation of drug concentration in the brain after lorlatinib administration.

AUC:
Area under the curve AUMC: Area under the moment curve ESI: Electron spray ionization FDA: Food and Drug Administration HPLC: High-performance liquid chromatography IS: Internal standard LLOQ: Lower limit of quantification LC-MS/MS: Liquid chromatography-tandem mass spectrometry MRT: Mean residence time NSCLC: Non-small cell lung cancer QC: Quality control TKIs: Tyrosine kinase inhibitors.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.