|Home | About | Journals | Submit | Contact Us | Français|
A sensitive and rapid liquid chromatography-tandem mass spectrometry method has been developed for the assessment of therapeutic exposures of aprepitant in HIV-infected patients and rhesus macaques. The method utilized a simple sample preparation procedure of protein precipitation with methanol. Chromatographic separation was performed on a reversed phase C8 column (Hypersil Gold, 50 × 2.1 mm, 3 μ) using a mobile phase composed of acetonitrile and water in 0.5% formic acid through gradient elution. Electro-spray ionization in positive mode was incorporated in the tandem mass spectrometric detection. The lower limit of quantitation of aprepitant in plasma of rhesus macaques & human and cerebral spinal fluid of rhesus macaques were 1, 1, and 0.1 ng/mL, respectively. The method has been successfully employed for the measurement of aprepitant in preclinical and clinical samples collected from 3 SIV-infected rhesus macaques and 10 patients with HIV infection. In conclusion, this liquid chromatography-tandem mass spectrometry method is suitable for preclinical-clinical translational research exploring exposure-response relationships with aprepitant as well as therapeutic drug monitoring of aprepitant.
Substance P (SP) is the most abundant neurokinin in mammalian CNS and a potent modulator of neuroimmunoregulation1. It has been reported that SP enhances HIV infection by directly assisting virus replication in macrophages and CD4+ cells and/or indirectly influencing HIV proliferation via induction of inflammatory cytokines (e.g., IL-1, IL-6, and TNF-α)1. NK-1R, one of the three human neurokinin receptors, is mainly responsible for mediating the biological responses of SP1. The NK-1R antagonist, aprepitant, was reported to down-regulate CCR5 receptor expression on monocyte-derived macrophages (MDM), and inhibit HIV R5 strain replication in MDM. NK-1R receptor antagonists might also reverse the impairment of NK cell function found in HIV infection via antagonism again substance P, whose effects are mediated through NK-1R receptor2.
Aprepitant (Emend®), a neurokinin-1 receptor (NK-1R) antagonist, is licensed by the United States FDA as an antiemetic against chemotherapy induced emesis and marketed by Merck & Co. in 2003. Currently, aprepitant is also under evaluation as a new therapy in NeuroAIDS patients from the Integrated Preclinical and Clinical Program (IPCP) grant mechanism supported by the NIH at the Children’s Hospital of Philadelphia and University of Pennsylvania2,3. Development of sensitive bioanalytical methods to detect the exposure of aprepitant and its metabolites in biological fluids (e.g., plasma, cerebral spinal fluid [CSF]), is crucially important to facilitate pharmacokinetics and pharmacodynamics study in cell culture, simian immunodeficiency virus (SIV) infected rhesus macaques, and HIV-infected patients. Quantitation of aprepitant in human plasma has been reported using liquid-liquid extraction and HPLC-MS/MS with atmospheric-pressure chemical ionization (APCI) mass spectrometric detection4,5. In both of the published methods, the lower limit of quantitation (LLOQ) of aprepitant was reported as 10 ng/mL in human plasma4,5. In order to better characterize aprepitant in rhesus macaque CSF and human plasma, we incorporated protein precipitation and HPLC-MS/MS with electrospray ionization technique to develop a more sensitive and rapid bioanalytical method to quantify aprepitant in CSF and plasma. Our assay was validated in the concentration range of 0.1-10 ng/mL in CSF and 1-1,000 ng/mL in plasma of rhesus macaque and human, respectively. Compared with traditional aprepitant sample preparation procedures4,5, including liquid-liquid extraction, nitrogen blowing-down, and reconstitution with mobile phase, time duration and efforts for sample preparation in our assays was dramatically reduced by using simple protein precipitation method, which is more suitable for preparing infectious samples from HIV-infected patients in hospital and other clinical research laboratories. This method can be applied to therapeutic drug monitoring of aprepitant in clinics.
The purpose of this research was to develop a sensitive, time- and cost-efficient HPLC-MS/MS method to determine concentrations of aprepitant in CSF and plasma of rhesus macaque and human. This is the first report describing bioanalytical methods for aprepitant detection in CSF. The method has been applied to pharmacokinetics/pharmacodynamics (PK/PD) study of aprepitant in SIV-infected rhesus macaques and HIV-infected patients.
Aprepitant was received from the Merck Research Laboratories (Rahway, NJ, USA). Internal standard (IS), quadrideuterated aprepitant (aprepitant-d4) was purchased from SynFine Research Inc (Ontario, Canada). HPLC grade acetonitrile and methanol (CHROMASOLV®) was purchased from Sigma-Aldrich (St. Louis, MO, USA). Formic acid (A.C.S. reagent grade, Rieldel-de Haën®) was obtained from Sigma-Aldrich Fluka (St. Louis, MO, USA). OmniSolv HPLC grade water was purchased from (EMD Chemicals Inc, Gibbstown, NJ, USA). Rhesus macaque CSF and plasma were purchased from Bioreclamation Inc (East Meadow, NY, USA). Blank human plasma was obtained from blood bank at the Children’s Hospital of Philadelphia (CHOP) (Philadelphia, PA, USA).
Sample analysis was performed on an API 3000 mass spectrometer (Applied Biosystems/MDS Sciex, Toronto, Canada) coupled with a Shimadzu HPLC system (Shimadzu, Columbia, MD, USA) using electrospray ionization (ESI). The Shimadzu HPLC system consists of two LC-10ADVP delivery pumps, a DGU-14A vacuum degasser, and a SIL-HTC autosampler. The Shimadzu system is connected with API 3000 via a 6-port Valco valve (VICI Valco Instruments Co. Inc, Houston, TX, USA).
Chromatographic separation was conducted on a Hypersil Gold C8 column (50 × 2.1 mm, 3 μm) with a Hypersil Gold C8 guard column (10 × 2.1 mm, 3 μm) (Thermo Electron Corporation, Waltham, MA, USA). Compounds of interest were separated from interference using a gradient mobile phase comprised of 0.5% formic acid water (A) and acetonitrile with 0.5% formic acid (B) at a flow rate of 0.3 mL/min. The mobile phase was comprised of a 90:10 (v/v) mixture of components A and B for the first 1.5 min of each chromatographic run, increased to 98% of B in a linear gradient from 1.5 to 2.5 min, kept at 98% of B till 4.2 min, and then returned to 10% of B at 4.3 min. The equilibration time for the column with the initial mobile phase was 2.9 min. The Valco valve was programmed to divert HPLC flow to waste when data acquisition was not required.
Mass spectrometric detection was performed using an ESI source in positive mode under the following conditions: curtain gas, 12 units; nebulizer gas flow, 12 psi; turboIonSpray gas flow, 7-8 L/min; collision gas, 6 units; turboIonSpray (IS) voltage, 3,500 V; entrance potential (EP), 10 V; collision energy (CE), 29 V; source temperature, 500 °C; and dwell time, 200 ms. The optimized declustering potential (DP) and collision cell exit potential (CXP), were set at 40 and 15 V, respectively. Multiple reaction monitoring (MRM) was used to detect aprepitant and IS at 535.3/277.1 and 539.3/281.1, respectively. Analytical data were acquired and integrated by Analyst software (version 1.4.2; Applied Biosystems/MDS Sciex, Toronto, Canada).
Stock standard solutions of aprepitant and internal standard (1 mg/mL) were prepared in methanol. The stock standard solution of aprepitant was diluted with methanol to yield a 100 μg/mL stock solution. This solution was further diluted with solution of methanol to give a series of aprepitant working standards of 0.02 to 20 μg/mL and QC standards of 0.02, 0.05, 2, 8, and 16 μg/mL for plasma of rhesus macaque and human. Aprepitant CSF standards ranged from 2 to 200 ng/mL and QC standards contained 2, 5, 20, and 160 ng/mL for rhesus macaque. IS solution was with 1% formic acid in methanol made at 100 ng/mL for plasma samples and 10 ng/mL for CSF samples.
Different concentrations of working solutions of standards for calibration curves and QC standards were added to blank plasma or CSF to give different sets of plasma or CSF samples for calibration curves and QC standards. The calibration curve for plasma was constructed with 8 standards of aprepitant at 1, 2, 5, 25, 125, 375, 500, and 1000 ng/mL in plasma. QC standards for plasma contained 4 concentrations of aprepitant at 2.5, 100, 400, and 800 ng/mL in plasma. For CSF samples of rhesus macaque, calibration curves was built with 6 standards at 0.1, 0.2, 0.5, 1.25, 5, and 10 ng/mL; QC standards were made at 0.25, 1, and 8 ng/mL.
The rhesus macaque blood and CSF samples were collected from 3 SIV-infected rhesus macaques in Tulane National Primate Research Center (Covington, LA, USA). Blood samples were drawn at 0, 1, 2, 4, 8, and 12 hr on day 1, 7, and 14, respectively, when macaques were orally administrated with 80 mg- or 125 mg- capsules of aprepitant (Emend®, Merck & Co., Inc., Whitehouse Station, NJ, USA) daily. CSF samples were collected at trough level on day 1, 7, and 14 before oral administration of aprepitant in rhesus macaques.
Human blood samples were collected from HIV-infected patients recruited with age no less than 18 years old at School of Medicine University of Pennsylvania (Philadelphia, PA, USA). Human blood samples were collected at 0, 0.5, 1, 2, 4, and 8 hr, respectively, on day 1 and 14 following capsule doses of 125 mg- or 250 mg-aprepitant was orally administered to each HIV-infected patient daily. Trough level blood was obtained on day 3, 7, and 10 as well for all the patients. Generally, blood and CSF samples were stored in the tube containing heparin as the anticoagulant, then centrifuged (within 2 hours from collection) at 1,800 g for 15 minutes at 4 °C. The supernatant was then transferred to polypropylene tubes and stored at -70 °C until LC-MS/MS analysis.
One hundred micro liters of plasma standards (standard curve), quality control standards, and/or rhesus macaque or human samples were added into individual 2-mL centrifuge tubes (VWR, West Chester, PA, USA). After 300 μL of IS solution was added into each tube except tubes containing blank plasma and methanol with 1% formic acid, tubes were capped and vortexed for 3 min, then centrifuged at 17,390 g for 10 min. Three hundred micro liters of supernatant was transferred into an HPLC insert for LC-MS/MS analysis. Five micro liters of supernatant was injected into the LC-MS/MS system.
Fifty micro liters of CSF standards (standard curve), quality control standards, and/or rhesus macaque samples were added into individual 2-mL centrifuge tubes (VWR, West Chester, PA, USA). After 150 μL of IS solution was added into each tube except tubes containing blank plasma and methanol with 1% formic acid, the same procedures were followed as described in the plasma sample preparation. Finally, 150 μL of supernatant was transferred into an HPLC insert for LC-MS/MS analysis. Thirty micro liters of supernatant was injected into the LC-MS/MS system.
Method validation was conducted in accordance with bioanalytical method validation guidelines for industry enacted by U.S. FDA6. Blank biological matrices were extracted without fortification of analyte and IS to determine the extent to which endogenous substances are comprised of the interference at the retention time and precursor/fragment ion values of the analyte and IS. Chromatograms were evaluated by a unique combination of retention time, precursor, and fragment ions for both the analyte and IS. The limit of detection (LOD) is typically determined to be in the region where the signal to noise ratio is greater than 5. QC standards and LLOQ in both plasma and CSF, were subjected to preparation procedures described above, and injected into LC-MS/MS. The assays described above were repeated five times within the same day to obtain intraday precision and over 3 different days to obtain inter-day precision, both expressed as a percentage of relative standard deviation values (RSD%).
Calibration curves were constructed with corresponding sets of standards in biological matrix (plasma or CSF) described in Preparation of Standards for Calibration Curves and QC Standards in Biological Matrix. QC standards were run with calibration curves to ensure the quality of sample analysis. Sample preparation and LC-MS/MS analysis were performed in triplicate for each data point. The analyte/IS peak area ratios obtained were plotted against the corresponding concentrations of the analytes. Calibration curves were constructed using a weighted 1/x2 linear regression. The values of LLOQ were calculated, according to FDA guidelines of bioanalytical method validation for industry, as the analyte response should be at least 5 times the response compared to blank response6. QC standards in corresponding biological matrix were used to calculate accuracy according to the following equation: 100 × [predicted concentration−nominal concentration]/nominal concentration.
Due to high lipophilicity of aprepitant, C8 HPLC columns was chosen instead of commonly selected C18 columns where it takes much higher percentage of organic solution and longer time to elute aprepitant and thereby incurs coelution with endogenous substances. Additionally, symmetric peaks are displayed for the basic compound of aprepitant on Hypersil columns, compared with tailing peaks exhibited in several other brands of columns chosen during the method development.
Two functions are involved in IS solution: one is to add IS to correct errors occurred during the process of extraction and sample injection on LC-MS/MS; the other is to extract aprepitant out of biological matrix with maximum recovery and precipitate protein. Addition of formic acid in the IS solution facilitated the extraction of aprepitant and IS with high recovery rate from the biological matrix due to the basic characteristics of aprepitant. Percentage of formic acid added in IS solution has been tested and 1% formic acid served the best of extraction rate without compromising the integrity of aprepitant compound in the biological matrix and processed samples for LC-MS/MS analysis. Compared with liquid-liquid extraction procedures for aprepitant human plasma samples published4,5, the protein precipitation procedure we applied dramatically reduced time, equipments and materials, and human power in this otherwise traditionally time- and effort- consuming part of bioanalysis. In addition, this simple sample preparation procedure provides greater benefits in terms of safety and efficiency when preparing clinical samples from HIV-infected patients.
The peak areas of aprepitant at the LLOQ were at least five times greater than those of interference substances in all three biological matrices examined. No interference of aprepitant was observed in blank rhesus macaque CSF. The LOD was 0.05, 0.5, and 0.5 ng/mL for aprepitant in rhesus macaque CSF, rhesus macaque plasma, and human plasma, respectively. Calibration curves were set up in three biological matrices. Good linearity (r2> 0.9962) was found in calibration ranges of aprepitant in five different lots of all three biological matrices, demonstrating linearity over the entire standard curve range. The LLOQ was 0.1, 1, and 1 ng/mL for aprepitant in rhesus macaque CSF, rhesus macaque plasma, and human plasma, respectively (Figure 1, ,2,2, & 3). Typical equations for the calibration curves for aprepitant was y = 0.0455x + 0.000714 (rhesus macaque CSF), y = 0.00414x + 0.000645 (rhesus macaque plasma), y = 0.0041x + 0.00043 (human plasma), respectively.
Accuracy and precision assays were carried out using LLOQ and QC standards. The results of these assays are given in Table 1. The RSD values of precision assays were lower than 13.2%.
The potential for carryover effect was investigated by injecting a sequence of the least two successive aliquots of extracted plasma/CSF samples with corresponding extracted samples of the highest calibration concentration (i.e., 1000 ng/mL for plasma; 100 ng/mL for CSF) of aprepitant in standard curves into LC-MS/MS system followed by at least three successive aliquots of extracted drug free plasma/CSF sample. The carryover effect was calculated as less than 0.1% of the highest calibration concentration in corresponding biological matrix.
The potential for matrix effects was tested by comparing the peak area of aprepitant and IS from plasma/CSF samples spiked before protein precipitation procedure and neat solution among the different sources of plasma/CSF samples. Absolute recovery with the range of 96--102% was established in plasma and CSF, when comparing difference between peak areas of both analyte and IS and /or peak area ratios of the samples spiked before protein precipitation and those of neat solutions. In addition, the matrix effect was not observed as indicated by small coefficient of variation (< 5.5%) of the slopes of the calibration curves in different lots of plasma and CSF (Table 2). As such, the protein precipitation procedure coupled with suitable chromatographic conditions ensured that there was no matrix effect among different lots of plasma/CSF.
There is no significant difference in assay concentrations for processed samples from an analytical run, including non-zero standards, a blank, and a control, and all QC standards, after stored in the HPLC autosampler set at 4 °C or in a refrigerator for at least 24 hr. No significant difference in aprepitant concentrations was observed for the rhesus macaque plasma and CSF samples of aprepitant stored in -70 °C for two years and human plasma samples stored in -70 °C for one year.
The validated method has been utilized in a PK/PD studies in rhesus macaques and HIV-infected patients. The analysis of plasma and CSF samples from a rhesus macaque and a patient with HIV infection are given in Figures 4 and and55 respectively. The method continues to provide reliable data in an ongoing Phase IB PK/PD/safety trial in Neuro-AIDS patients. We expect to use these results to evaluate aprepitant exposure-response characteristics based on its immunologic, virologic and CNS-mediated antipsychotic effects. Allometric scaling to facilitate preclinical to clinical bridging and the animal disease model and dose optimization modeling to guide future clinical investigation are ongoing efforts in our laboratory reliant on this data.
The LC-MS/MS method developed here is sensitive and rapid using a simple sample preparation—protein precipitation. When compared with the results obtained from previous published LC-MS/MS analyses with aprepitant, the present method allows the determination of aprepitant at lower concentrations (LLOQ = 1 ng/mL instead of LLOQ =10 ng/mL in plasma; LLOQ = 0.1 ng/mL in CSF, first time reported). In addition, our sample preparation procedure of simple protein precipitation, if compared with published sample pretreatment procedure using solid phase extraction procedure for aprepitant, demonstrated better results in terms of precision and extraction yield with lower consumption of organic solvent / materials and equipment, time, and human capital. Furthermore, a small volume of plasma (100 μL) or CSF (50 μL) was required for this method, thus reducing the amount of the blood and CSF and minimizing collection difficulties in HIV-infected patients and SIV-infected rhesus macaques, especially with CSF sample collection.
In conclusion, this method has been shown to have good precision, high accuracy, and satisfactory stability for aprepitant detection in biological matrices. It is well suited for cell culture, PK/PD and metabolism study of aprepitant in animals and humans. Due to its simplicity, accuracy, and efficiency, this method can be well applied to clinical settings where therapeutic drug monitoring in patients treated with aprepitant is warranted.
We thank Dr. Pyone Pyone Aye for providing plasma and CSF samples obtained from control and SIV-infected rhesus macaques in aprepitant preclinical study at the Tulane National Primate Research Center, Dr. Pablo Tebas for supporting us with human plasma samples obtained from HIV-infected patients being treated with aprepitant in a phase IB, placebo controlled, double blind trial held at University of Pennsylvania School of Medicine, Dr. Florin Tuluc for kindly providing aprepitant standards for this analysis work. This work was supported by NIH Grant, P01 MH076388.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.