ORIGINAL ARTICLE
Development and Validation of an HPLC-UV Method for Sorafenib Quantification in Human Plasma and Application to Patients With Cancer in Routine Clinical Practice Vanesa Escudero-Ortiz, PhD,*‡ Juan José Pérez-Ruixo, PhD,† and Belén Valenzuela, PhD*‡
Background: Several factors such as low therapeutic index, large interindividual variability in systemic exposure, and the relationships between exposure and toxicity for sorafenib could justify its therapeutic drug monitoring (TDM). To support TDM, a selective and precise high-performance liquid chromatography with ultraviolet detection (HPLC-UV) method was developed and validated for the determination of sorafenib in human plasma.
Methods: After protein precipitation with acetonitrile, sorafenib and lapatinib (internal standard) were separated using isocratic elution on a Kromasil C18 column using a mobile phase of acetonitrile and 20 mmol/L ammonium acetate in a proportion 53:47 (vol/vol) pumped at a constant flow rate of 1.2 mL/min. Quantification was performed at 260 nm. Validation experiments were carried out after the guidelines for Bioanalytical Method Validation published by the Food and Drug Administration and the European Medicines Agency. Results: Calibration curves were linear over the range 0.1–20 mcg/mL. Inter- and intra-day coefficients of variation were ,3%. The limit of detection and the lower limit of quantification were 0.06 and 0.1 mcg/mL, respectively. Recoveries of sorafenib from plasma were .99% in all cases. Conclusions: This method was successfully applied to the determination of the drug in the plasma of 2 patients with cancer receiving sorafenib 200 and 400 mg orally twice daily, respectively, and could be useful for TDM of sorafenib in routine clinical practice. Key Words: sorafenib, HPLC-UV, cancer, therapeutic drug monitoring, interindividual variability (Ther Drug Monit 2014;36:317–325)
Received for publication May 8, 2013; accepted October 22, 2013. From the *Platform of Oncology, Hospital Quirón Torrevieja, Torrevieja; †Pharmacokinetics and Drug Metabolism, AMGEN, Valencia; and ‡Cathedra of Multidisciplinary Oncology, UCAM Universidad San Antonion de Murcia, Spain. The authors declare no conflict of interest. The views expressed in this article are the personal views of the authors reflecting their scientific knowledge of this topic and should not be understood or quoted as being made on behalf of the companies for which the authors currently work. Correspondence: Belén Valenzuela, PhD, Platform of Oncology, Hospital Quirón Torrevieja, Ptda. de la Loma, s/n, Torrevieja 03180, Spain (e-mail: [email protected]). Copyright © 2013 by Lippincott Williams & Wilkins
Ther Drug Monit Volume 36, Number 3, June 2014
INTRODUCTION Sorafenib (Nexavar) is a potent competitive oral multikinase inhibitor that inhibits the Raf serine/threonine kinases pathway (Raf-1, wild-type B-Raf, and b-raf V600E), the vascular endothelial growth factor receptor, and also intracellular members of the mitogen-activated protein kinase signal transduction pathway.1 Sorafenib is approved for the treatment of patients with advanced renal cell carcinoma and unresectable hepatocellular carcinoma.2,3 The recommended dose for adults is 400 mg twice daily (bid) on a continuous dosing schedule and should be administered at least 1 hour before or 2 hours after food intake.2,3 Although absolute bioavailability of sorafenib has not been reported, probably because the drug only exists in oral form, it is well known that concomitant food intake reduces the bioavailability by 29% in patients with cancer.4 In this setting, the median time to reach peak concentration (tmax) is approximately 4 hours (range, 1–12.3 hours). At the recommended sorafenib dose, the maximal plasma concentration (Cmax) was in the range of 2.3–3.0 mcg/mL after the first dose and increased up to 5.4–10.0 mcg/mL at steady-state condition, which is reached after 7 days1 because the terminal halflife (t1/2) varies between 25 and 48 hours.5 In different phase I trials,1 a less than proportional increase in the area under the curve (AUC) and Cmax was observed with increasing doses of sorafenib, and more recently, Hornecker et al,5 bioavailability decrease at high daily doses, resulting from a saturation of the intestinal absorption. In fact, dividing the daily dose in 3 doses seems to reduce the saturation of the absorption leading to higher exposure than bid dosing. In vitro studies with human plasma show that around 99.8% was bound to plasma proteins, mainly to albumin, and to a lesser extent, to a1-acid glycoprotein, and the other low density lipoproteins.6 As a consequence of high lipophilicity and plasma protein binding, sorafenib presents a high apparent steady-state volume of distribution (Vss/F) of 213 L.7 Sorafenib is metabolized primarily in the liver to an N-oxide metabolite by cytochrome P450 (CYP) 3A4 isoform (,5%) and glucuronidation-mediated uridine diphosphate glucuronosyl transferase 1A9 isoform (around 15%).8,9 After oral administration of sorafenib, 77% was excreted in the feces (50% as parent drug), and 19% was excreted in the urine mainly as glucuronide conjugates.9 Sorafenib undergoes enterohepatic circulation with typical double peaks in the plasma concentration–time profiles in patients treated with sorafenib.7
317
Escudero-Ortiz et al
As for many tyrosine kinases inhibitors, sorafenib also shows a large interindividual variability in the absorption, distribution, metabolism, and excretion processes.10 Sorafenib plasma exposure is positively correlated with the likelihood of toxicity in heavily pretreated patients with advanced, metastatic, or recurrent solid tumors, so that pharmacokinetic (PK) variability contributes to the variability in drug exposure and therefore in the variation of pharmacodynamic effects.10 Actually, Boudou-Rouquette et al,11 reported that patients who experienced grade 3–4 toxicities, such as hand–foot skin reactions, asthenia, and diarrhea, had a median of AUC0–12 greater than that observed in the remaining patients (61.9 vs 53.0 mg$h/L, P = 0.017). Therapeutic drug monitoring (TDM) has proved to be useful for drugs with a clear exposure–toxicity relationship and moderate to higher PK variability, as in the case of sorafenib.10 In fact, TDM could be useful to prevent severe toxicities that could result in a discontinuation of sorafenib therapy and detect patients with suboptimal sorafenib exposure before disease progression. Most of analytical methods described in the literature to quantify sorafenib plasma concentrations used high-performance liquid chromatography with tandem mass spectrometric detection (HPLC-MS/MS).12–17 This equipment is expensive and not available in all clinical laboratories. Recently, 2 HPLC methods with ultraviolet (UV) detection method have been developed for the determination and quantification of sorafenib concentration in human plasma.18,19 In these analytical methods, the sample preparation is performed by a liquid–liquid extraction, which involves both more time in the bioassay and the use of organic solvents. Therefore, in this study, a simple, rapid, and sensitive HPLC-UV method was developed and validated for the measurement of sorafenib in human plasma using lapatinib as the internal standard (IS). This simple and rapid method uses a protein precipitation for sample preparation that requires less time than liquid–liquid extraction and less expenses of organic solvents. Potentially, this assay could be applied to sorafenib TDM in routine clinical practice.
MATERIALS AND METHODS Chemicals and Reagents Sorafenib, p-toluenesulfonate salt (,99% purity, lot BSF105), and lapatinib di-p-toluenesulfonate salt (,99% purity, lot BLP-108) were obtained from LC Laboratories (Woburn, MA). HPLC-grade acetonitrile and ammonium acetate were obtained from Panreac Química S. A. (Barcelona, Spain). Dimethylsulfoxide was obtained from Sigma–Aldrich Química (Madrid, Spain). Purified deionized water was obtained with a HydroReverse osmosis system connected to the Milli-Q UV Plus purifying system (Millipore, Milford, MA). Drug-free plasma used for the preparation of quality controls (QCs) and calibrators was kindly provided from the Centro de Transfusiones de la Comunidad Valenciana (San Juan, Alicante, Spain).
Chromatographic Conditions Chromatographic analysis was performed using an HPLC system (Agilent 1200 serie) equipped with a degasser
318
Ther Drug Monit Volume 36, Number 3, June 2014
(model G1233A), quaternary pump (model G1354A), autosampler (model G1329A), thermostated column compartment (model G1316A), and an UV-visible detector (model G1365B). Data were acquired and processed with HP Chem Station chromatography manager software from Agilent Technologies (Santa Clara, CA). Separation of the compounds of interest was achieved using a Kromasil C18 column (5 mm; 4.6 · 150 mm) with a guard column packed with the same bonded phase (5 mm; 4.6 · 10 mm). The chromatographic separation was carried out using a mobile phase (consisting of a mixture of 0.02 mol/L ammonium acetate and acetonitrile; 47:53; vol/vol) pumped at a constant flow rate of 1.2 mL/min. The column was maintained at 258C, and the eluents were monitored at a wavelength of 260 nm.
Stock Solutions Stock solutions containing sorafenib and lapatinib (5000 mcg/mL) were prepared in dimethylsulfoxide and were stored at 2208C in the dark. Each day, working solutions of sorafenib were prepared freshly by diluting the stock solutions with acetonitrile to obtain concentrations of 5, 25, 50, and 500 mcg/mL. Likewise, working solutions of the IS were further diluted with acetonitrile to a final concentration of 1.5 mcg/mL.
Preparation of Standard Solutions To validate the analytical method, calibrators and QC samples were prepared. Calibrators were plasma samples containing known concentrations of the analyte of interest (sorafenib). They were independently prepared in the matrix mimicking as much as possible the future routine analysis of sorafenib samples. These calibrators were used to construct a calibration curve consisting of a blank sample (matrix sample processed without IS), a zero sample (matrix sample processed with IS), and 8 nonzero samples covering the expected range, including the lower limit of quantification (LLOQ). Sorafenib calibrators were prepared by diluting the working solutions further with blank human plasma each day to obtain concentrations of 0.1, 0.5, 1.5, 4.0, 6.0, 9.0, 15, and 20 mcg/mL. QC samples were of known concentrations of analyte in human plasma. These controls spanned the calibration curve, encompassing concentrations at 0.5, 4.0, and 15 mcg/mL. Both calibrators and QC samples were analyzed in the same way as patient plasma samples.
Sample Preparation Frozen samples were thawed at room temperature. A 200-mL aliquot of calibrator, QC, or patient sample was pipetted into a polypropylene microcentrifuge tube, where 700 mL of acetonitrile containing the IS was added. The tube was vortex mixed for 10 seconds, followed by centrifugation at 8900g for 10 minutes at room temperature. The supernatant was transferred into a glass tube and evaporated to dryness under vacuum (out of the light). The residue was reconstituted with 150 mL of the HPLC mobile phase, and after vortex mixing for 20 seconds, the clear supernatant was transferred 2013 Lippincott Williams & Wilkins
Ther Drug Monit Volume 36, Number 3, June 2014
to tinted microvials, and the autosampler was programmed to inject 100 mL into the HPLC system.
Method Validation The method validation was carried out following the guidelines for Bioanalytical Method Validation published by the Food and Drug Administration20 and the European Medicines Agency.21
Linearity, Limit of Detection, and Lower Limit of Quantification Complete calibration curves (8 concentrations ranging from 0.1 to 20 mcg/mL) were analyzed on 3 separate days. Calibration curves were constructed as peak height ratios of sorafenib to lapatinib, and then they were plotted against the nominal concentration of each sorafenib concentration. Linear regression analysis of the calibration data was performed using the equation: y ¼ a þ b · C; where y is the peak height ratio, C is the nominal concentration of sorafenib in calibration samples, and a and b are the intercept and the slope of the curve, respectively. The model homoscedasticity was assessed by the Levene test. Best weighting factor for linear regression was determined according to the result of the Levene test and the evolution of variance with respect to concentration as has been published previously.18 Slope, intercept, and correlation coefficient (r) were calculated for each calibration curve. The assay sensitivity was evaluated by determining the limit of detection (LOD) and the LLOQ. LOD was defined as the concentration of analyte required to give a signal equal to the blank plus 3 times the SD of the blank, whereas LLOQ was the lowest concentration of analyte required to give a signal equal to the blank plus 10 times the SD of the blank and acceptable accuracy and imprecision data. Both parameters were determined empirically by analysis of a series of decreasing concentrations of enriched plasma samples in multiple replicates. The following conditions had to be met in developing the calibration curve: 1. ,20% deviation of the LLOQ from the nominal concentration. 2. ,15% deviation of calibrators, other than LLOQ, from the nominal concentration. 3. At least two-thirds of the calibrators should meet the criteria given in 1 and 2. In addition, linearity in the calibration range from 0.1 to 20 mcg/mL was concluded, if r was greater than 0.99 for all 3 calibration curves.
Recovery Sorafenib recovery was evaluated at concentrations corresponding to the QCs (0.5, 4.0, and 15 mcg/mL). Recovery after extraction was determined by comparing the peak area of 2013 Lippincott Williams & Wilkins
HPLC-UV Method for Sorafenib Quantification
the extracted plasma with that of the identical concentration of analyte prepared in the mobile phase without extraction, representing 100% recovery. For each QC, 5 replicates were analyzed. Analyte recovery did not necessarily need to be 100%, but the extent of analyte recovery and IS should be consistent, precise, and reproducible along the concentrations of the calibration curve.
Selectivity Blank human plasma samples from 6 different patients were prepared as described above to check for peaks that might interfere with the detection of the analyte and/or the IS. A zero sample was analyzed to check for the absence of interference with the analyte. The interference peak should be ,5% of the peak area for the LLOQ for both the analyte and IS in plasma. Blood sampling was approved by the local Ethics Committee, and patients were informed about the risk of the blood extraction and provided written informed consent before the procedure.
Inaccuracy and Imprecision Inaccuracy and imprecision were assessed by determining sorafenib concentrations at LLOQ and in QC samples at 0.5, 4.0, 15 mcg/mL, measuring 5 replicates per concentration on 3 different days. The inaccuracy was expressed by the mean relative error (MRE) and calculated as the difference between the nominal concentration (Ct) and the mean of the observed concentration (Cobs) as follows: Inaccuracy ¼
Ct 2 Cobs · 100 Ct
The imprecision was expressed as relative standard deviation (RSD) of the different measurements for each nominal concentration using the equation: Imprecision ¼
SDCobs · 100; Cobs
where SDCobs represents the SD of Cobs and Cobs represent the mean of Cobs. For each concentration, both MRE and RSD should be ,15% except for LLOQ, where they should not deviate by .20%.
Stability Stock and working solutions stability, freeze–thaw stability, short-term light stability, long-term stability, and autosampler stability studies were performed. The results of the stability tests were expressed as inaccuracy and imprecision as reported in the previous section. For all stability studies, the solution was considered as stable if the difference between these percentages was not .15%, as indicated in the guidelines for bioanalytical methods validation.22 Stock and Working Solutions The stock solutions of sorafenib and lapatinib were stored in the dark at 2708C and compared after 21 days with respect to a freshly made stock solution. A maximum
319
Escudero-Ortiz et al
Ther Drug Monit Volume 36, Number 3, June 2014
FIGURE 1. Representative chromatograms: blank human plasma without sorafenib (A), blank human plasma at the LLOQ (0.1 mcg/mL) (B), blank human plasma after the addition of 20 mcg/mL of sorafenib (C), and a plasma sample from a cancer patient treated with 400 mg bid of sorafenib (plasma concentration determined was equal to 6.1 mcg/mL) (D). IS, internal standard; mAU, milli-absorbance units; min, minutes.
320
2013 Lippincott Williams & Wilkins
Ther Drug Monit Volume 36, Number 3, June 2014
HPLC-UV Method for Sorafenib Quantification
FIGURE 2. Representative chromatograms of 6 plasma samples from 6 patients with cancer. IS, internal standard; mAU, milliabsorbance units; min, minutes.
duration of 21 days was selected as the maximum time the frozen stock solutions were kept. The stability of working solutions of sorafenib and lapatinib, prepared freshly each day, was evaluated after 8 hours at room temperature with and without light. 2013 Lippincott Williams & Wilkins
Freeze and Thaw Stability Freeze–thaw stability of sorafenib was determined by assaying the 3 QCs in triplicate over 3 freeze–thaw cycles. Three aliquots at each concentration were stored at 2708C for 24 hours and thawed unassisted at room temperature. When
321
Ther Drug Monit Volume 36, Number 3, June 2014
Escudero-Ortiz et al
TABLE 1. Parameters of the Calibration Curves Linearity Equation y = a + b · C Day Day Day Day Day Day
1 2 3 4 5 6
a (·1022)
b (·1025)
Correlation Coefficient, r
4.6 3.1 3.1 3.7 6.8 5.6
2.3 2.4 2.3 2.4 2.2 2.3
0.999 0.999 0.999 1.000 0.999 0.998
completely thawed, the samples were refrozen for 24 hours under the same conditions. The freeze–thaw cycle was repeated 2 more times. Short-Term Light Stability The stability of sorafenib in plasma at room temperature with and without light was investigated by comparing, in triplicate, the 3 QCs stored in these conditions to freshly extracted samples. Four aliquots of each concentration were stored at room temperature for 1, 2, 4, and 8 hours. A maximum duration of 8 hours was selected to coincide with regular daily working time. Long-Term Stability The stability of sorafenib in plasma at 2708C was evaluated by assaying, in triplicate, the 3 QCs samples stored at this temperature for 21 days. No more than 21 days was evaluated, as this is our maximal storage condition for patient samples. Autosampler Stability Sorafenib and lapatinib stabilities after plasma extraction were evaluated by keeping the 3 extracted QC samples in triplicate in the autosampler at room temperature for 2 and 10 hours. About 10 hours was selected because it is the time required to analyze 50 samples, the maximum number of samples that can be processed within a working day in 1 system.
Sorafenib TDM The assay was used to quantify the steady-state sorafenib plasma concentrations from 2 patients with cancer. To be eligible, patients had to have been treated with sorafenib for at least 7 days, the time necessary to reach the sorafenib steady state. Blood samples (5 mL) were collected
in lithium heparin tubes before drug administration (that is representative of plasma concentration at the end of the interval, ie, 12 hours) and at 1, 3, 6, and 8 hours after drug administration. The tube was wrapped in aluminum foil to totally protect it against light and centrifuged at 1000g for 10 minutes at room temperature. Plasma supernatant was separated and stored at 2208C until analysis. Patients were informed about the risk and benefits of repeated blood sampling for PK studies and provided written informed consent before the treatment. Blood sampling for TDM was approved by the local Ethics Committee as a routine procedure for these patients during chemotherapy. A maximum a posteriori estimation method of the sorafenib individual PK parameters was implemented in the NONMEM VI level 2.0 software package (ICON, Hanover, MD) using the POSTHOC option.23 The results of a previous population PK analysis were used as a prior information to describe the time course of sorafenib plasma concentration. The PK model is a 1-compartment model with 4 transit absorption compartments, enterohepatic circulation, and the firstorder elimination kinetics.7 Graphical analyses were performed using S-Plus 6.1 Professional Edition (Insightful, Seattle, WA).
RESULTS Typical chromatograms obtained with extracted drugfree human plasma, samples of plasma spiked with sorafenib (0.1 and 20 mcg/mL) and IS, and plasma from 1 patient treated with sorafenib (400 mg bid) and spiked with IS were clean and without interferences (Fig. 1). Retention times for sorafenib and lapatinib were 9.8 and 2.7 minutes, respectively. The assay was found to be specific and selective because no interference was observed with biological compounds in 6 plasma samples from 6 patients with cancer. The pharmacological treatments used in these patients involved other anticancer drugs, statins, oral antidiabetics, antihypertensives, diuretics, antidepressants, antibiotics, and anesthetics because these drugs could be regularly administered with sorafenib or lapatinib. The list of the drugs coadministered in each patient is provided in Figure 2. For linearity assessment, the Levene statistic test showed a significant difference (P , 0.05) between variance for each concentration of the calibration samples. As the variance grew more than proportionally to the concentration, the best weighting factor was 1 per peak area ratio.24 Calibration
TABLE 2. Inaccuracy and Precision of Sorafenib Determination in Human Plasma Theoretical Concentration (mcg/mL) 0.1 0.5 4.0 15
Intraday* Cobs (SD) (mcg/mL) 0.10 0.48 4.06 15.64
Interday*
Inaccuracy (MRE, %)
Precision (RSD, %)
21.2 4.8 21.5 24.2
2.8 2.4 1.2 4.0
(0.01) (0.01) (0.05) (0.63)
Cobs (SD) (mcg/mL) 0.10 0.47 3.97 15.76
(0.01) (0.01) (0.10) (0.43)
Inaccuracy (MRE, %)
Precision (RSD, %)
21.6 5.6 0.8 25.1
2.7 2.4 2.6 2.7
*Results are expressed as mean (SD) from 5 replicates. Cobs, observed concentration.
322
2013 Lippincott Williams & Wilkins
Ther Drug Monit Volume 36, Number 3, June 2014
HPLC-UV Method for Sorafenib Quantification
TABLE 3. Assessment of Sorafenib Stability in Human Plasma Theoretical Concentration (mcg/mL) 0.5 4.0 15
Freeze–Thaw Stability
Short-term Stability With Light
Short-term Stability Without Light
Cycle 1
Cycle 3
1h
8h
1h
8h
Long-term Stability 21 d
26.0 (0.6) 0.6 (0.6) 22.1 (3.9)
22.9 (1.9) 24.9 (3.2) 210.7 (3.1)
22.7 (1.2) 3.5 (2.0) 3.7 (4.1)
9.9 (6.1) 5.1 (7.5) 5.6 (0.6)
2.8 (1.3) 1.6 (6.2) 5.6 (6.1)
12.4 (1.7) 210.8 (2.8) 213.2 (2.7)
22.2 (2.1) 15.0 (3.3) 3.8 (1.9)
Results are expressed as inaccuracy value (coefficient of variation of replicates, CV%). Inaccuracy was calculated as 100 · (measured concentration 2 theoretical concentration)/(theoretical concentration).
parameters are shown in Table 1, and r was .0.998 on the 3 validation days. The LLOQ was the lowest concentration of the calibration curves (0.1 mcg/mL), and the LOD was determined to be 0.06 mcg/mL. In addition, 100% of the evaluated samples for the LLOQ showed MRE and RSE values that did not deviate .20%. For all other concentrations tested, all samples (100%) showed a deviation from the nominal value of ,15% (data not shown). These data confirm that the assay meets the acceptance criteria in relation to the linearity specified by the Food and Drug Administration and the European Medicines Agency. The absolute recoveries of sorafenib from plasma at concentrations of 0.5, 4.0, and 15 mcg/mL were 101.4 6 2.3%, 99.5 6 0.5%, and 102.5 6 2.1%, respectively. The absolute recovery of the IS was 100.6 6 2.3%. These values showed that the recovery of the analyte after protein precipitation was reproducible across the concentrations of the calibration curve according to the criteria specified by regulatory agencies. Intra- and inter-assay inaccuracy and imprecision are reported in Table 2. For the 3 QC concentrations, intra- and inter-assay inaccuracy ranged from 21.5% to 4.8% and from 0.8% to 5.6%, respectively. RSD for the intra- and inter-assay results was lower than 4.1% and 2.8%, respectively. Results
of MRE and RSD for the LLOQ were ,2.9%. Low values of intra- and inter-day bias and imprecision showed that assay is reliable and reproducible relative to the requirements of regulatory agencies. The stock solutions of sorafenib and lapatinib stored for 21 days at 2708C were comparable to the freshly made ones, the value for inaccuracy (coefficient of variation for the replicates CV%) obtained was ,14.8% (4.9%) for both drugs. Working solutions of sorafenib and lapatinib prepared daily were stable for at least 8 hours at room temperature with or without light, the value for inaccuracy (CV%) obtained was ,10.8% (0.4%) for both drugs. Table 3 shows the main results of additional stability studies performed for plasma samples. Data for the cycle 2 of freeze–thaw stability and data for 2 and 8 hours of short-term stability with and without light are not shown in this table, but the inaccuracy value was below 13.3% in all samples for these stability studies. Sorafenib was also stable up to 10 hours on the autosampler without any significant degradation (,13.2%). Figure 3 shows the time course of plasma concentration sorafenib for 2 patients, 1 received 200 mg bid (patient A) of sorafenib and another patient received 400 mg bid (patient B). Both patients showed good tolerance to the drug, and there were
FIGURE 3. Sorafenib plasma concentration–time profile in patients after an oral dose of 200 mg bid (patient A) or 400 mg bid (patient B). The ordinate axis represents, on logarithmic scale, the plasma concentration of sorafenib. Sorafenib plasma concentrations observed (points); 25th, 50th, and 75th percentiles (solid light gray lines); model-based 50% prediction interval (gray shaded area). AUC0–12, area under the curve from 0 to 12 hours; Cl/F, apparent clearance of the drug from plasma; Vss/F, apparent steady-state volume of distribution. 2013 Lippincott Williams & Wilkins
323
Ther Drug Monit Volume 36, Number 3, June 2014
Escudero-Ortiz et al
no adverse effects reported. Bayesian prediction (solid green line) and individual PK parameters are shown in Figure 3. As can be seen, steady-state (trough) concentrations were detectable and higher than the LLOQ of the analytical method. Moreover, in both patients, there is an increase in plasma concentrations of sorafenib after 6 hours following administration of a dose. This increase corresponds to the typical double peak in the plasma concentration–time profiles in patients treated with sorafenib because of the enterohepatic circulation.7 The assay was capable of measuring the plasma concentrations of sorafenib in patients undergoing sorafenib therapy.
DISCUSSION This assay was developed and validated for the measurement of sorafenib in human plasma so that it could be used in routine clinical practice for sorafenib TDM in patients receiving this drug orally. During the assay optimization phase, different stationary and mobile phase compositions, such as the mixture of potassium dihydrogen phosphate buffer and acetonitrile or methanol, were evaluated for their capability to separate sorafenib, lapatinib, and endogenous compounds. The mixture already described of 0.02 mol/L ammonium acetate buffer and acetonitrile as organic modifier in a proportion 47:53, vol/vol; produced optimal separation with acceptable retention time and very sharp and symmetrical peak shapes for both the sorafenib and lapatinib as reported elsewhere.25 Total run time for each sample analysis was 12 minutes, which is approximately 15% less than the run time previously reported for other HPLC-UV assays.18 The choice of the IS is a critical aspect of the method development because it influences both accuracy and imprecision, which are particularly important aspects in HPLC-UV analytical methods. Lapatinib was chosen as IS because its chemical structures and physiochemical properties are very similar to sorafenib, it had a shorter elution time than sorafenib, and it eluted in a chromatogram area where no matrix interferences were detected. In addition, lapatinib is not used in combination with sorafenib in clinical settings. Furthermore, during the process of assay development and validation, it is necessary to ensure the stability of the solutions. In this regard, the stability studies performed showed that stock and working solutions, plasma samples, and reconstituted samples after the plasma extraction process were stable under the working conditions described to analyze plasma samples of patients treated with sorafenib. These results are fairly comparable to other sorafenib stability studies published previously.12,19,20 The LLOQ was similar to the previously published values with other HPLC-UV assay19 and 14-fold higher than previously reported with an LC-MS/MS method.16 However, 2 phase I studies reported that Cmin, ss sorafenib concentrations in patients receiving 200 mg bid was close to 1.2 mcg/mL, which is 12-fold higher than the LLOQ.26,27 As the standard dose of sorafenib is around 400 mg bid, the assay LLOQ (0.1 mcg/mL) is deemed adequate to monitor the sorafenib concentrations in patients with cancer.
324
Until now, only 2 HPLC-UV assays were published for the determination of sorafenib in human plasma.18,19 Both assays used sample preparation by liquid–liquid extraction that, in contrast to the protein precipitation extraction described in this article, involves more time and a risk for the technician because of the use of organic solvents. In addition, the absolute recoveries of sorafenib from plasma obtained in the present analytical method were .99% in the entire concentration range evaluated. This value was higher than those previously reported by Blanchet el al18 with a liquid–liquid extraction (absolute recoveries of sorafenib .68%). In conclusion, the HPLC assay described here is sensitive, selective, and reproducible and can be used for the accurate and precise determination of sorafenib in human plasma for PK studies and TDM. UV detection provides the required level of sensitivity for measuring pharmacologically relevant concentration of sorafenib in patients with cancer, and it has been successfully implemented in routine clinical practice for sorafenib TDM. This analytical method provides a readily available, practicable, and robust method to monitor sorafenib concentrations in patients with cancer, and it may also contribute to the spreading of sorafenib TDM.
ACKNOWLEDGMENTS The authors thank the patients, medical and nursing staff of the Hospital Quirón Torrevieja who participated in this study. REFERENCES 1. Strumberg D, Clark JW, Awada A, et al. Safety, pharmacokinetics, and preliminary antitumor activity of sorafenib: a review of four phase I trials in patients with advanced refractory solid tumors. Oncologist. 2007;12: 426–437. 2. Sorafenib summary of product characteristics. Available at: http://www. ema.europa.eu/docs/en_GB/document_library/EPAR Summary_for_the_ public/human/000690/WC500027705.pdf. Accessed April 10, 2013. 3. Nexavar (sorafenib) label information. Available at: http://www. accessdata.fda.gov/drugsatfda_docs/label/2010/021923s008s009lbl.pdf. Accessed April 10, 2013. 4. Kane RC, Farrell AT, Saber H, et al. Sorafenib for the treatment of advanced renal cell carcinoma. Clin Cancer Res. 2006;12:7271–7278. 5. Hornecker M, Blanchet B, Billemont B, et al. Saturable absorption of sorafenib in patients with solid tumors: a population model. Invest New Drugs. 2012;30:1991–2000. 6. Villarroel MC, Pratz KW, Xu L, et al. Plasma protein binding of sorafenib, a multi kinase inhibitor: in vitro and in cancer patients. Invest New Drugs. 2012;30:2096–2102. 7. Jain L, Woo S, Gardner ER, et al. Population pharmacokinetic analysis of sorafenib in patients with solid tumours. Br J Clin Pharmacol. 2011;72: 294–305. 8. van Erp NP, Gelderblom H, Guchelaar HJ. Clinical pharmacokinetics of tyrosine kinase inhibitors. Cancer Treat Rev. 2009;35:692–706. 9. Lathia C, Lettieri J, Cihon F, et al. Lack of effect of ketoconazolemediated CYP3A4 inhibition on sorafenib clinical pharmacokinetics. Cancer Chemother Pharmacol. 2006;57:685–692. 10. Klümpen HJ, Samer CF, Mathijssen RH, et al. Moving towards dose individualization of tyrosine kinase inhibitors. Cancer Treat Rev. 2011; 37:251–260. 11. Boudou-Rouquette P, Ropert S, Mir O, et al. Variability of sorafenib toxicity and exposure over time: a pharmacokinetic/pharmacodynamic analysis. Oncologist. 2012;17:1204–1212. 12. Zhao M, Rudek MA, He P, et al. A rapid and sensitive method for determination of sorafenib in human plasma using a liquid chromatog-
2013 Lippincott Williams & Wilkins
Ther Drug Monit Volume 36, Number 3, June 2014
13. 14.
15. 16.
17.
18.
19.
raphy/tandem mass spectrometry assay. J Chromatogr B Analyt Technol Biomed Life Sci. 2007;846:1–7. Jain L, Gardner ER, Venitz J, et al. Development of a rapid and sensitive LC-MS/MS assay for the determination of sorafenib in human plasma. J Pharm Biomed Anal. 2008;46:362–367. Sparidans RW, Vlaming ML, Lagas JS, et al. Liquid chromatographytandem mass spectrometric assay for sorafenib and sorafenib-glucuronide in mouse plasma and liver homogenate and identification of the glucuronide metabolite. J Chromatogr B Analyt Technol Biomed Life Sci. 2009;877:269–276. Bouchet S, Chauzit E, Ducint D, et al. Simultaneous determination of nine tyrosine kinase inhibitors by 96-well solid-phase extraction and ultra performance LC/MS-MS. Clin Chim Acta. 2011;412:1060–1067. Andriamanana I, Gana I, Duretz B, et al. Simultaneous analysis of anticancer agents bortezomib, imatinib, nilotinib, dasatinib, erlotinib, lapatinib, sorafenib, sunitinib and vandetanib in human plasma using LC/MS/ MS. J Chromatogr B Analyt Technol Biomed Life Sci. 2013;926:83–91. Lankheet NA, Hillebrand MJ, Rosing H, et al. Method development and validation for the quantification of dasatinib, erlotinib, gefitinib, imatinib, lapatinib, nilotinib, sorafenib and sunitinib in human plasma by liquid chromatography coupled with tandem mass spectrometry. Biomed Chromatogr. 2013;27:466–476. Blanchet B, Billemont B, Cramard J, et al. Validation of an HPLC-UV method for sorafenib determination in human plasma and application to cancer patients in routine clinical practice. J Pharm Biomed Anal. 2009; 49:1109–1114. Heinz WJ, Kahle K, Helle-Beyersdorf A, et al. High-performance liquid chromatographic method for the determination of sorafenib in human serum and peritoneal fluid. Cancer Chemother Pharmacol. 2011;68: 239–245.
2013 Lippincott Williams & Wilkins
HPLC-UV Method for Sorafenib Quantification
20. Guidance for Industry. Bioanalytical method validation. U.S. department of health and human services. Food and Drug Administration center for drug evaluation and research (CDER). Available at: http://www.fda.gov/ downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidan ces/ ucm070107.pdf. Accessed December 17, 2012. 21. Validation of analytical procedures: text and methodology. Ich topic Q2 (R1). CPMP/ICH/381/95-ICH Q2 (R1). Available at: http://www.ich.org/ products/guidelines/quality/quality-single/article/validation-of-analyticalprocedures-text-and-methodology.html. Accessed November 27, 2013. 22. Guideline on bioanalytical method validation. EMA/CHMP/EWP/192217/ 2009. Available at: http://www.ema.europa.eu/docs/en_GB/document_ library/Scientific_guideline/2011/08/WC500109686.pdf. Accessed February 17, 2013. 23. Beal SL, Sheiner LB, Boeckman AJ, eds. NONMEN Users Guides (1989–2006). Ellicott, MA: ICON Development Solutions; 1989. 24. Rusnak DW, Lackey K, Affleck K, et al. The effects of the novel, reversible epidermal growth factor receptor/ErbB-2 tyrosine kinase inhibitor, GW2016, on the growth of human normal and tumor-derived cell lines in vitro and in vivo. Mol Cancer Ther. 2001;1:85–94. 25. Escudero-Ortiz V, Pérez-Ruixo JJ, Valenzuela B. Development and validation of a simple and sensitive HPLC-UV method for lapatinib quantification in human plasma. Ther Drug Monit. 2013;35:796–802. 26. Nabors LB, Supko JG, Rosenfeld M, et al. Phase I trial of sorafenib in patients with recurrent or progressive malignant glioma. Neuro Oncol. 2011;13:1324–1330. 27. Navid F, Baker SD, McCarville MB, et al. Phase I and clinical pharmacology study of bevacizumab, sorafenib, and low-dose cyclophosphamide in children and young adults with refractory/recurrent solid tumors. Clin Cancer Res. 2013;19:236–246.
325
Development and Validation of an HPLC-UV Method for Sorafenib Quantification in Human Plasma and Application to Patients With Cancer in Routine Clinical Practice Vanesa Escudero-Ortiz, PhD,*‡ Juan José Pérez-Ruixo, PhD,† and Belén Valenzuela, PhD*‡
Background: Several factors such as low therapeutic index, large interindividual variability in systemic exposure, and the relationships between exposure and toxicity for sorafenib could justify its therapeutic drug monitoring (TDM). To support TDM, a selective and precise high-performance liquid chromatography with ultraviolet detection (HPLC-UV) method was developed and validated for the determination of sorafenib in human plasma.
Methods: After protein precipitation with acetonitrile, sorafenib and lapatinib (internal standard) were separated using isocratic elution on a Kromasil C18 column using a mobile phase of acetonitrile and 20 mmol/L ammonium acetate in a proportion 53:47 (vol/vol) pumped at a constant flow rate of 1.2 mL/min. Quantification was performed at 260 nm. Validation experiments were carried out after the guidelines for Bioanalytical Method Validation published by the Food and Drug Administration and the European Medicines Agency. Results: Calibration curves were linear over the range 0.1–20 mcg/mL. Inter- and intra-day coefficients of variation were ,3%. The limit of detection and the lower limit of quantification were 0.06 and 0.1 mcg/mL, respectively. Recoveries of sorafenib from plasma were .99% in all cases. Conclusions: This method was successfully applied to the determination of the drug in the plasma of 2 patients with cancer receiving sorafenib 200 and 400 mg orally twice daily, respectively, and could be useful for TDM of sorafenib in routine clinical practice. Key Words: sorafenib, HPLC-UV, cancer, therapeutic drug monitoring, interindividual variability (Ther Drug Monit 2014;36:317–325)
Received for publication May 8, 2013; accepted October 22, 2013. From the *Platform of Oncology, Hospital Quirón Torrevieja, Torrevieja; †Pharmacokinetics and Drug Metabolism, AMGEN, Valencia; and ‡Cathedra of Multidisciplinary Oncology, UCAM Universidad San Antonion de Murcia, Spain. The authors declare no conflict of interest. The views expressed in this article are the personal views of the authors reflecting their scientific knowledge of this topic and should not be understood or quoted as being made on behalf of the companies for which the authors currently work. Correspondence: Belén Valenzuela, PhD, Platform of Oncology, Hospital Quirón Torrevieja, Ptda. de la Loma, s/n, Torrevieja 03180, Spain (e-mail: [email protected]). Copyright © 2013 by Lippincott Williams & Wilkins
Ther Drug Monit Volume 36, Number 3, June 2014
INTRODUCTION Sorafenib (Nexavar) is a potent competitive oral multikinase inhibitor that inhibits the Raf serine/threonine kinases pathway (Raf-1, wild-type B-Raf, and b-raf V600E), the vascular endothelial growth factor receptor, and also intracellular members of the mitogen-activated protein kinase signal transduction pathway.1 Sorafenib is approved for the treatment of patients with advanced renal cell carcinoma and unresectable hepatocellular carcinoma.2,3 The recommended dose for adults is 400 mg twice daily (bid) on a continuous dosing schedule and should be administered at least 1 hour before or 2 hours after food intake.2,3 Although absolute bioavailability of sorafenib has not been reported, probably because the drug only exists in oral form, it is well known that concomitant food intake reduces the bioavailability by 29% in patients with cancer.4 In this setting, the median time to reach peak concentration (tmax) is approximately 4 hours (range, 1–12.3 hours). At the recommended sorafenib dose, the maximal plasma concentration (Cmax) was in the range of 2.3–3.0 mcg/mL after the first dose and increased up to 5.4–10.0 mcg/mL at steady-state condition, which is reached after 7 days1 because the terminal halflife (t1/2) varies between 25 and 48 hours.5 In different phase I trials,1 a less than proportional increase in the area under the curve (AUC) and Cmax was observed with increasing doses of sorafenib, and more recently, Hornecker et al,5 bioavailability decrease at high daily doses, resulting from a saturation of the intestinal absorption. In fact, dividing the daily dose in 3 doses seems to reduce the saturation of the absorption leading to higher exposure than bid dosing. In vitro studies with human plasma show that around 99.8% was bound to plasma proteins, mainly to albumin, and to a lesser extent, to a1-acid glycoprotein, and the other low density lipoproteins.6 As a consequence of high lipophilicity and plasma protein binding, sorafenib presents a high apparent steady-state volume of distribution (Vss/F) of 213 L.7 Sorafenib is metabolized primarily in the liver to an N-oxide metabolite by cytochrome P450 (CYP) 3A4 isoform (,5%) and glucuronidation-mediated uridine diphosphate glucuronosyl transferase 1A9 isoform (around 15%).8,9 After oral administration of sorafenib, 77% was excreted in the feces (50% as parent drug), and 19% was excreted in the urine mainly as glucuronide conjugates.9 Sorafenib undergoes enterohepatic circulation with typical double peaks in the plasma concentration–time profiles in patients treated with sorafenib.7
317
Escudero-Ortiz et al
As for many tyrosine kinases inhibitors, sorafenib also shows a large interindividual variability in the absorption, distribution, metabolism, and excretion processes.10 Sorafenib plasma exposure is positively correlated with the likelihood of toxicity in heavily pretreated patients with advanced, metastatic, or recurrent solid tumors, so that pharmacokinetic (PK) variability contributes to the variability in drug exposure and therefore in the variation of pharmacodynamic effects.10 Actually, Boudou-Rouquette et al,11 reported that patients who experienced grade 3–4 toxicities, such as hand–foot skin reactions, asthenia, and diarrhea, had a median of AUC0–12 greater than that observed in the remaining patients (61.9 vs 53.0 mg$h/L, P = 0.017). Therapeutic drug monitoring (TDM) has proved to be useful for drugs with a clear exposure–toxicity relationship and moderate to higher PK variability, as in the case of sorafenib.10 In fact, TDM could be useful to prevent severe toxicities that could result in a discontinuation of sorafenib therapy and detect patients with suboptimal sorafenib exposure before disease progression. Most of analytical methods described in the literature to quantify sorafenib plasma concentrations used high-performance liquid chromatography with tandem mass spectrometric detection (HPLC-MS/MS).12–17 This equipment is expensive and not available in all clinical laboratories. Recently, 2 HPLC methods with ultraviolet (UV) detection method have been developed for the determination and quantification of sorafenib concentration in human plasma.18,19 In these analytical methods, the sample preparation is performed by a liquid–liquid extraction, which involves both more time in the bioassay and the use of organic solvents. Therefore, in this study, a simple, rapid, and sensitive HPLC-UV method was developed and validated for the measurement of sorafenib in human plasma using lapatinib as the internal standard (IS). This simple and rapid method uses a protein precipitation for sample preparation that requires less time than liquid–liquid extraction and less expenses of organic solvents. Potentially, this assay could be applied to sorafenib TDM in routine clinical practice.
MATERIALS AND METHODS Chemicals and Reagents Sorafenib, p-toluenesulfonate salt (,99% purity, lot BSF105), and lapatinib di-p-toluenesulfonate salt (,99% purity, lot BLP-108) were obtained from LC Laboratories (Woburn, MA). HPLC-grade acetonitrile and ammonium acetate were obtained from Panreac Química S. A. (Barcelona, Spain). Dimethylsulfoxide was obtained from Sigma–Aldrich Química (Madrid, Spain). Purified deionized water was obtained with a HydroReverse osmosis system connected to the Milli-Q UV Plus purifying system (Millipore, Milford, MA). Drug-free plasma used for the preparation of quality controls (QCs) and calibrators was kindly provided from the Centro de Transfusiones de la Comunidad Valenciana (San Juan, Alicante, Spain).
Chromatographic Conditions Chromatographic analysis was performed using an HPLC system (Agilent 1200 serie) equipped with a degasser
318
Ther Drug Monit Volume 36, Number 3, June 2014
(model G1233A), quaternary pump (model G1354A), autosampler (model G1329A), thermostated column compartment (model G1316A), and an UV-visible detector (model G1365B). Data were acquired and processed with HP Chem Station chromatography manager software from Agilent Technologies (Santa Clara, CA). Separation of the compounds of interest was achieved using a Kromasil C18 column (5 mm; 4.6 · 150 mm) with a guard column packed with the same bonded phase (5 mm; 4.6 · 10 mm). The chromatographic separation was carried out using a mobile phase (consisting of a mixture of 0.02 mol/L ammonium acetate and acetonitrile; 47:53; vol/vol) pumped at a constant flow rate of 1.2 mL/min. The column was maintained at 258C, and the eluents were monitored at a wavelength of 260 nm.
Stock Solutions Stock solutions containing sorafenib and lapatinib (5000 mcg/mL) were prepared in dimethylsulfoxide and were stored at 2208C in the dark. Each day, working solutions of sorafenib were prepared freshly by diluting the stock solutions with acetonitrile to obtain concentrations of 5, 25, 50, and 500 mcg/mL. Likewise, working solutions of the IS were further diluted with acetonitrile to a final concentration of 1.5 mcg/mL.
Preparation of Standard Solutions To validate the analytical method, calibrators and QC samples were prepared. Calibrators were plasma samples containing known concentrations of the analyte of interest (sorafenib). They were independently prepared in the matrix mimicking as much as possible the future routine analysis of sorafenib samples. These calibrators were used to construct a calibration curve consisting of a blank sample (matrix sample processed without IS), a zero sample (matrix sample processed with IS), and 8 nonzero samples covering the expected range, including the lower limit of quantification (LLOQ). Sorafenib calibrators were prepared by diluting the working solutions further with blank human plasma each day to obtain concentrations of 0.1, 0.5, 1.5, 4.0, 6.0, 9.0, 15, and 20 mcg/mL. QC samples were of known concentrations of analyte in human plasma. These controls spanned the calibration curve, encompassing concentrations at 0.5, 4.0, and 15 mcg/mL. Both calibrators and QC samples were analyzed in the same way as patient plasma samples.
Sample Preparation Frozen samples were thawed at room temperature. A 200-mL aliquot of calibrator, QC, or patient sample was pipetted into a polypropylene microcentrifuge tube, where 700 mL of acetonitrile containing the IS was added. The tube was vortex mixed for 10 seconds, followed by centrifugation at 8900g for 10 minutes at room temperature. The supernatant was transferred into a glass tube and evaporated to dryness under vacuum (out of the light). The residue was reconstituted with 150 mL of the HPLC mobile phase, and after vortex mixing for 20 seconds, the clear supernatant was transferred 2013 Lippincott Williams & Wilkins
Ther Drug Monit Volume 36, Number 3, June 2014
to tinted microvials, and the autosampler was programmed to inject 100 mL into the HPLC system.
Method Validation The method validation was carried out following the guidelines for Bioanalytical Method Validation published by the Food and Drug Administration20 and the European Medicines Agency.21
Linearity, Limit of Detection, and Lower Limit of Quantification Complete calibration curves (8 concentrations ranging from 0.1 to 20 mcg/mL) were analyzed on 3 separate days. Calibration curves were constructed as peak height ratios of sorafenib to lapatinib, and then they were plotted against the nominal concentration of each sorafenib concentration. Linear regression analysis of the calibration data was performed using the equation: y ¼ a þ b · C; where y is the peak height ratio, C is the nominal concentration of sorafenib in calibration samples, and a and b are the intercept and the slope of the curve, respectively. The model homoscedasticity was assessed by the Levene test. Best weighting factor for linear regression was determined according to the result of the Levene test and the evolution of variance with respect to concentration as has been published previously.18 Slope, intercept, and correlation coefficient (r) were calculated for each calibration curve. The assay sensitivity was evaluated by determining the limit of detection (LOD) and the LLOQ. LOD was defined as the concentration of analyte required to give a signal equal to the blank plus 3 times the SD of the blank, whereas LLOQ was the lowest concentration of analyte required to give a signal equal to the blank plus 10 times the SD of the blank and acceptable accuracy and imprecision data. Both parameters were determined empirically by analysis of a series of decreasing concentrations of enriched plasma samples in multiple replicates. The following conditions had to be met in developing the calibration curve: 1. ,20% deviation of the LLOQ from the nominal concentration. 2. ,15% deviation of calibrators, other than LLOQ, from the nominal concentration. 3. At least two-thirds of the calibrators should meet the criteria given in 1 and 2. In addition, linearity in the calibration range from 0.1 to 20 mcg/mL was concluded, if r was greater than 0.99 for all 3 calibration curves.
Recovery Sorafenib recovery was evaluated at concentrations corresponding to the QCs (0.5, 4.0, and 15 mcg/mL). Recovery after extraction was determined by comparing the peak area of 2013 Lippincott Williams & Wilkins
HPLC-UV Method for Sorafenib Quantification
the extracted plasma with that of the identical concentration of analyte prepared in the mobile phase without extraction, representing 100% recovery. For each QC, 5 replicates were analyzed. Analyte recovery did not necessarily need to be 100%, but the extent of analyte recovery and IS should be consistent, precise, and reproducible along the concentrations of the calibration curve.
Selectivity Blank human plasma samples from 6 different patients were prepared as described above to check for peaks that might interfere with the detection of the analyte and/or the IS. A zero sample was analyzed to check for the absence of interference with the analyte. The interference peak should be ,5% of the peak area for the LLOQ for both the analyte and IS in plasma. Blood sampling was approved by the local Ethics Committee, and patients were informed about the risk of the blood extraction and provided written informed consent before the procedure.
Inaccuracy and Imprecision Inaccuracy and imprecision were assessed by determining sorafenib concentrations at LLOQ and in QC samples at 0.5, 4.0, 15 mcg/mL, measuring 5 replicates per concentration on 3 different days. The inaccuracy was expressed by the mean relative error (MRE) and calculated as the difference between the nominal concentration (Ct) and the mean of the observed concentration (Cobs) as follows: Inaccuracy ¼
Ct 2 Cobs · 100 Ct
The imprecision was expressed as relative standard deviation (RSD) of the different measurements for each nominal concentration using the equation: Imprecision ¼
SDCobs · 100; Cobs
where SDCobs represents the SD of Cobs and Cobs represent the mean of Cobs. For each concentration, both MRE and RSD should be ,15% except for LLOQ, where they should not deviate by .20%.
Stability Stock and working solutions stability, freeze–thaw stability, short-term light stability, long-term stability, and autosampler stability studies were performed. The results of the stability tests were expressed as inaccuracy and imprecision as reported in the previous section. For all stability studies, the solution was considered as stable if the difference between these percentages was not .15%, as indicated in the guidelines for bioanalytical methods validation.22 Stock and Working Solutions The stock solutions of sorafenib and lapatinib were stored in the dark at 2708C and compared after 21 days with respect to a freshly made stock solution. A maximum
319
Escudero-Ortiz et al
Ther Drug Monit Volume 36, Number 3, June 2014
FIGURE 1. Representative chromatograms: blank human plasma without sorafenib (A), blank human plasma at the LLOQ (0.1 mcg/mL) (B), blank human plasma after the addition of 20 mcg/mL of sorafenib (C), and a plasma sample from a cancer patient treated with 400 mg bid of sorafenib (plasma concentration determined was equal to 6.1 mcg/mL) (D). IS, internal standard; mAU, milli-absorbance units; min, minutes.
320
2013 Lippincott Williams & Wilkins
Ther Drug Monit Volume 36, Number 3, June 2014
HPLC-UV Method for Sorafenib Quantification
FIGURE 2. Representative chromatograms of 6 plasma samples from 6 patients with cancer. IS, internal standard; mAU, milliabsorbance units; min, minutes.
duration of 21 days was selected as the maximum time the frozen stock solutions were kept. The stability of working solutions of sorafenib and lapatinib, prepared freshly each day, was evaluated after 8 hours at room temperature with and without light. 2013 Lippincott Williams & Wilkins
Freeze and Thaw Stability Freeze–thaw stability of sorafenib was determined by assaying the 3 QCs in triplicate over 3 freeze–thaw cycles. Three aliquots at each concentration were stored at 2708C for 24 hours and thawed unassisted at room temperature. When
321
Ther Drug Monit Volume 36, Number 3, June 2014
Escudero-Ortiz et al
TABLE 1. Parameters of the Calibration Curves Linearity Equation y = a + b · C Day Day Day Day Day Day
1 2 3 4 5 6
a (·1022)
b (·1025)
Correlation Coefficient, r
4.6 3.1 3.1 3.7 6.8 5.6
2.3 2.4 2.3 2.4 2.2 2.3
0.999 0.999 0.999 1.000 0.999 0.998
completely thawed, the samples were refrozen for 24 hours under the same conditions. The freeze–thaw cycle was repeated 2 more times. Short-Term Light Stability The stability of sorafenib in plasma at room temperature with and without light was investigated by comparing, in triplicate, the 3 QCs stored in these conditions to freshly extracted samples. Four aliquots of each concentration were stored at room temperature for 1, 2, 4, and 8 hours. A maximum duration of 8 hours was selected to coincide with regular daily working time. Long-Term Stability The stability of sorafenib in plasma at 2708C was evaluated by assaying, in triplicate, the 3 QCs samples stored at this temperature for 21 days. No more than 21 days was evaluated, as this is our maximal storage condition for patient samples. Autosampler Stability Sorafenib and lapatinib stabilities after plasma extraction were evaluated by keeping the 3 extracted QC samples in triplicate in the autosampler at room temperature for 2 and 10 hours. About 10 hours was selected because it is the time required to analyze 50 samples, the maximum number of samples that can be processed within a working day in 1 system.
Sorafenib TDM The assay was used to quantify the steady-state sorafenib plasma concentrations from 2 patients with cancer. To be eligible, patients had to have been treated with sorafenib for at least 7 days, the time necessary to reach the sorafenib steady state. Blood samples (5 mL) were collected
in lithium heparin tubes before drug administration (that is representative of plasma concentration at the end of the interval, ie, 12 hours) and at 1, 3, 6, and 8 hours after drug administration. The tube was wrapped in aluminum foil to totally protect it against light and centrifuged at 1000g for 10 minutes at room temperature. Plasma supernatant was separated and stored at 2208C until analysis. Patients were informed about the risk and benefits of repeated blood sampling for PK studies and provided written informed consent before the treatment. Blood sampling for TDM was approved by the local Ethics Committee as a routine procedure for these patients during chemotherapy. A maximum a posteriori estimation method of the sorafenib individual PK parameters was implemented in the NONMEM VI level 2.0 software package (ICON, Hanover, MD) using the POSTHOC option.23 The results of a previous population PK analysis were used as a prior information to describe the time course of sorafenib plasma concentration. The PK model is a 1-compartment model with 4 transit absorption compartments, enterohepatic circulation, and the firstorder elimination kinetics.7 Graphical analyses were performed using S-Plus 6.1 Professional Edition (Insightful, Seattle, WA).
RESULTS Typical chromatograms obtained with extracted drugfree human plasma, samples of plasma spiked with sorafenib (0.1 and 20 mcg/mL) and IS, and plasma from 1 patient treated with sorafenib (400 mg bid) and spiked with IS were clean and without interferences (Fig. 1). Retention times for sorafenib and lapatinib were 9.8 and 2.7 minutes, respectively. The assay was found to be specific and selective because no interference was observed with biological compounds in 6 plasma samples from 6 patients with cancer. The pharmacological treatments used in these patients involved other anticancer drugs, statins, oral antidiabetics, antihypertensives, diuretics, antidepressants, antibiotics, and anesthetics because these drugs could be regularly administered with sorafenib or lapatinib. The list of the drugs coadministered in each patient is provided in Figure 2. For linearity assessment, the Levene statistic test showed a significant difference (P , 0.05) between variance for each concentration of the calibration samples. As the variance grew more than proportionally to the concentration, the best weighting factor was 1 per peak area ratio.24 Calibration
TABLE 2. Inaccuracy and Precision of Sorafenib Determination in Human Plasma Theoretical Concentration (mcg/mL) 0.1 0.5 4.0 15
Intraday* Cobs (SD) (mcg/mL) 0.10 0.48 4.06 15.64
Interday*
Inaccuracy (MRE, %)
Precision (RSD, %)
21.2 4.8 21.5 24.2
2.8 2.4 1.2 4.0
(0.01) (0.01) (0.05) (0.63)
Cobs (SD) (mcg/mL) 0.10 0.47 3.97 15.76
(0.01) (0.01) (0.10) (0.43)
Inaccuracy (MRE, %)
Precision (RSD, %)
21.6 5.6 0.8 25.1
2.7 2.4 2.6 2.7
*Results are expressed as mean (SD) from 5 replicates. Cobs, observed concentration.
322
2013 Lippincott Williams & Wilkins
Ther Drug Monit Volume 36, Number 3, June 2014
HPLC-UV Method for Sorafenib Quantification
TABLE 3. Assessment of Sorafenib Stability in Human Plasma Theoretical Concentration (mcg/mL) 0.5 4.0 15
Freeze–Thaw Stability
Short-term Stability With Light
Short-term Stability Without Light
Cycle 1
Cycle 3
1h
8h
1h
8h
Long-term Stability 21 d
26.0 (0.6) 0.6 (0.6) 22.1 (3.9)
22.9 (1.9) 24.9 (3.2) 210.7 (3.1)
22.7 (1.2) 3.5 (2.0) 3.7 (4.1)
9.9 (6.1) 5.1 (7.5) 5.6 (0.6)
2.8 (1.3) 1.6 (6.2) 5.6 (6.1)
12.4 (1.7) 210.8 (2.8) 213.2 (2.7)
22.2 (2.1) 15.0 (3.3) 3.8 (1.9)
Results are expressed as inaccuracy value (coefficient of variation of replicates, CV%). Inaccuracy was calculated as 100 · (measured concentration 2 theoretical concentration)/(theoretical concentration).
parameters are shown in Table 1, and r was .0.998 on the 3 validation days. The LLOQ was the lowest concentration of the calibration curves (0.1 mcg/mL), and the LOD was determined to be 0.06 mcg/mL. In addition, 100% of the evaluated samples for the LLOQ showed MRE and RSE values that did not deviate .20%. For all other concentrations tested, all samples (100%) showed a deviation from the nominal value of ,15% (data not shown). These data confirm that the assay meets the acceptance criteria in relation to the linearity specified by the Food and Drug Administration and the European Medicines Agency. The absolute recoveries of sorafenib from plasma at concentrations of 0.5, 4.0, and 15 mcg/mL were 101.4 6 2.3%, 99.5 6 0.5%, and 102.5 6 2.1%, respectively. The absolute recovery of the IS was 100.6 6 2.3%. These values showed that the recovery of the analyte after protein precipitation was reproducible across the concentrations of the calibration curve according to the criteria specified by regulatory agencies. Intra- and inter-assay inaccuracy and imprecision are reported in Table 2. For the 3 QC concentrations, intra- and inter-assay inaccuracy ranged from 21.5% to 4.8% and from 0.8% to 5.6%, respectively. RSD for the intra- and inter-assay results was lower than 4.1% and 2.8%, respectively. Results
of MRE and RSD for the LLOQ were ,2.9%. Low values of intra- and inter-day bias and imprecision showed that assay is reliable and reproducible relative to the requirements of regulatory agencies. The stock solutions of sorafenib and lapatinib stored for 21 days at 2708C were comparable to the freshly made ones, the value for inaccuracy (coefficient of variation for the replicates CV%) obtained was ,14.8% (4.9%) for both drugs. Working solutions of sorafenib and lapatinib prepared daily were stable for at least 8 hours at room temperature with or without light, the value for inaccuracy (CV%) obtained was ,10.8% (0.4%) for both drugs. Table 3 shows the main results of additional stability studies performed for plasma samples. Data for the cycle 2 of freeze–thaw stability and data for 2 and 8 hours of short-term stability with and without light are not shown in this table, but the inaccuracy value was below 13.3% in all samples for these stability studies. Sorafenib was also stable up to 10 hours on the autosampler without any significant degradation (,13.2%). Figure 3 shows the time course of plasma concentration sorafenib for 2 patients, 1 received 200 mg bid (patient A) of sorafenib and another patient received 400 mg bid (patient B). Both patients showed good tolerance to the drug, and there were
FIGURE 3. Sorafenib plasma concentration–time profile in patients after an oral dose of 200 mg bid (patient A) or 400 mg bid (patient B). The ordinate axis represents, on logarithmic scale, the plasma concentration of sorafenib. Sorafenib plasma concentrations observed (points); 25th, 50th, and 75th percentiles (solid light gray lines); model-based 50% prediction interval (gray shaded area). AUC0–12, area under the curve from 0 to 12 hours; Cl/F, apparent clearance of the drug from plasma; Vss/F, apparent steady-state volume of distribution. 2013 Lippincott Williams & Wilkins
323
Ther Drug Monit Volume 36, Number 3, June 2014
Escudero-Ortiz et al
no adverse effects reported. Bayesian prediction (solid green line) and individual PK parameters are shown in Figure 3. As can be seen, steady-state (trough) concentrations were detectable and higher than the LLOQ of the analytical method. Moreover, in both patients, there is an increase in plasma concentrations of sorafenib after 6 hours following administration of a dose. This increase corresponds to the typical double peak in the plasma concentration–time profiles in patients treated with sorafenib because of the enterohepatic circulation.7 The assay was capable of measuring the plasma concentrations of sorafenib in patients undergoing sorafenib therapy.
DISCUSSION This assay was developed and validated for the measurement of sorafenib in human plasma so that it could be used in routine clinical practice for sorafenib TDM in patients receiving this drug orally. During the assay optimization phase, different stationary and mobile phase compositions, such as the mixture of potassium dihydrogen phosphate buffer and acetonitrile or methanol, were evaluated for their capability to separate sorafenib, lapatinib, and endogenous compounds. The mixture already described of 0.02 mol/L ammonium acetate buffer and acetonitrile as organic modifier in a proportion 47:53, vol/vol; produced optimal separation with acceptable retention time and very sharp and symmetrical peak shapes for both the sorafenib and lapatinib as reported elsewhere.25 Total run time for each sample analysis was 12 minutes, which is approximately 15% less than the run time previously reported for other HPLC-UV assays.18 The choice of the IS is a critical aspect of the method development because it influences both accuracy and imprecision, which are particularly important aspects in HPLC-UV analytical methods. Lapatinib was chosen as IS because its chemical structures and physiochemical properties are very similar to sorafenib, it had a shorter elution time than sorafenib, and it eluted in a chromatogram area where no matrix interferences were detected. In addition, lapatinib is not used in combination with sorafenib in clinical settings. Furthermore, during the process of assay development and validation, it is necessary to ensure the stability of the solutions. In this regard, the stability studies performed showed that stock and working solutions, plasma samples, and reconstituted samples after the plasma extraction process were stable under the working conditions described to analyze plasma samples of patients treated with sorafenib. These results are fairly comparable to other sorafenib stability studies published previously.12,19,20 The LLOQ was similar to the previously published values with other HPLC-UV assay19 and 14-fold higher than previously reported with an LC-MS/MS method.16 However, 2 phase I studies reported that Cmin, ss sorafenib concentrations in patients receiving 200 mg bid was close to 1.2 mcg/mL, which is 12-fold higher than the LLOQ.26,27 As the standard dose of sorafenib is around 400 mg bid, the assay LLOQ (0.1 mcg/mL) is deemed adequate to monitor the sorafenib concentrations in patients with cancer.
324
Until now, only 2 HPLC-UV assays were published for the determination of sorafenib in human plasma.18,19 Both assays used sample preparation by liquid–liquid extraction that, in contrast to the protein precipitation extraction described in this article, involves more time and a risk for the technician because of the use of organic solvents. In addition, the absolute recoveries of sorafenib from plasma obtained in the present analytical method were .99% in the entire concentration range evaluated. This value was higher than those previously reported by Blanchet el al18 with a liquid–liquid extraction (absolute recoveries of sorafenib .68%). In conclusion, the HPLC assay described here is sensitive, selective, and reproducible and can be used for the accurate and precise determination of sorafenib in human plasma for PK studies and TDM. UV detection provides the required level of sensitivity for measuring pharmacologically relevant concentration of sorafenib in patients with cancer, and it has been successfully implemented in routine clinical practice for sorafenib TDM. This analytical method provides a readily available, practicable, and robust method to monitor sorafenib concentrations in patients with cancer, and it may also contribute to the spreading of sorafenib TDM.
ACKNOWLEDGMENTS The authors thank the patients, medical and nursing staff of the Hospital Quirón Torrevieja who participated in this study. REFERENCES 1. Strumberg D, Clark JW, Awada A, et al. Safety, pharmacokinetics, and preliminary antitumor activity of sorafenib: a review of four phase I trials in patients with advanced refractory solid tumors. Oncologist. 2007;12: 426–437. 2. Sorafenib summary of product characteristics. Available at: http://www. ema.europa.eu/docs/en_GB/document_library/EPAR Summary_for_the_ public/human/000690/WC500027705.pdf. Accessed April 10, 2013. 3. Nexavar (sorafenib) label information. Available at: http://www. accessdata.fda.gov/drugsatfda_docs/label/2010/021923s008s009lbl.pdf. Accessed April 10, 2013. 4. Kane RC, Farrell AT, Saber H, et al. Sorafenib for the treatment of advanced renal cell carcinoma. Clin Cancer Res. 2006;12:7271–7278. 5. Hornecker M, Blanchet B, Billemont B, et al. Saturable absorption of sorafenib in patients with solid tumors: a population model. Invest New Drugs. 2012;30:1991–2000. 6. Villarroel MC, Pratz KW, Xu L, et al. Plasma protein binding of sorafenib, a multi kinase inhibitor: in vitro and in cancer patients. Invest New Drugs. 2012;30:2096–2102. 7. Jain L, Woo S, Gardner ER, et al. Population pharmacokinetic analysis of sorafenib in patients with solid tumours. Br J Clin Pharmacol. 2011;72: 294–305. 8. van Erp NP, Gelderblom H, Guchelaar HJ. Clinical pharmacokinetics of tyrosine kinase inhibitors. Cancer Treat Rev. 2009;35:692–706. 9. Lathia C, Lettieri J, Cihon F, et al. Lack of effect of ketoconazolemediated CYP3A4 inhibition on sorafenib clinical pharmacokinetics. Cancer Chemother Pharmacol. 2006;57:685–692. 10. Klümpen HJ, Samer CF, Mathijssen RH, et al. Moving towards dose individualization of tyrosine kinase inhibitors. Cancer Treat Rev. 2011; 37:251–260. 11. Boudou-Rouquette P, Ropert S, Mir O, et al. Variability of sorafenib toxicity and exposure over time: a pharmacokinetic/pharmacodynamic analysis. Oncologist. 2012;17:1204–1212. 12. Zhao M, Rudek MA, He P, et al. A rapid and sensitive method for determination of sorafenib in human plasma using a liquid chromatog-
2013 Lippincott Williams & Wilkins
Ther Drug Monit Volume 36, Number 3, June 2014
13. 14.
15. 16.
17.
18.
19.
raphy/tandem mass spectrometry assay. J Chromatogr B Analyt Technol Biomed Life Sci. 2007;846:1–7. Jain L, Gardner ER, Venitz J, et al. Development of a rapid and sensitive LC-MS/MS assay for the determination of sorafenib in human plasma. J Pharm Biomed Anal. 2008;46:362–367. Sparidans RW, Vlaming ML, Lagas JS, et al. Liquid chromatographytandem mass spectrometric assay for sorafenib and sorafenib-glucuronide in mouse plasma and liver homogenate and identification of the glucuronide metabolite. J Chromatogr B Analyt Technol Biomed Life Sci. 2009;877:269–276. Bouchet S, Chauzit E, Ducint D, et al. Simultaneous determination of nine tyrosine kinase inhibitors by 96-well solid-phase extraction and ultra performance LC/MS-MS. Clin Chim Acta. 2011;412:1060–1067. Andriamanana I, Gana I, Duretz B, et al. Simultaneous analysis of anticancer agents bortezomib, imatinib, nilotinib, dasatinib, erlotinib, lapatinib, sorafenib, sunitinib and vandetanib in human plasma using LC/MS/ MS. J Chromatogr B Analyt Technol Biomed Life Sci. 2013;926:83–91. Lankheet NA, Hillebrand MJ, Rosing H, et al. Method development and validation for the quantification of dasatinib, erlotinib, gefitinib, imatinib, lapatinib, nilotinib, sorafenib and sunitinib in human plasma by liquid chromatography coupled with tandem mass spectrometry. Biomed Chromatogr. 2013;27:466–476. Blanchet B, Billemont B, Cramard J, et al. Validation of an HPLC-UV method for sorafenib determination in human plasma and application to cancer patients in routine clinical practice. J Pharm Biomed Anal. 2009; 49:1109–1114. Heinz WJ, Kahle K, Helle-Beyersdorf A, et al. High-performance liquid chromatographic method for the determination of sorafenib in human serum and peritoneal fluid. Cancer Chemother Pharmacol. 2011;68: 239–245.
2013 Lippincott Williams & Wilkins
HPLC-UV Method for Sorafenib Quantification
20. Guidance for Industry. Bioanalytical method validation. U.S. department of health and human services. Food and Drug Administration center for drug evaluation and research (CDER). Available at: http://www.fda.gov/ downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidan ces/ ucm070107.pdf. Accessed December 17, 2012. 21. Validation of analytical procedures: text and methodology. Ich topic Q2 (R1). CPMP/ICH/381/95-ICH Q2 (R1). Available at: http://www.ich.org/ products/guidelines/quality/quality-single/article/validation-of-analyticalprocedures-text-and-methodology.html. Accessed November 27, 2013. 22. Guideline on bioanalytical method validation. EMA/CHMP/EWP/192217/ 2009. Available at: http://www.ema.europa.eu/docs/en_GB/document_ library/Scientific_guideline/2011/08/WC500109686.pdf. Accessed February 17, 2013. 23. Beal SL, Sheiner LB, Boeckman AJ, eds. NONMEN Users Guides (1989–2006). Ellicott, MA: ICON Development Solutions; 1989. 24. Rusnak DW, Lackey K, Affleck K, et al. The effects of the novel, reversible epidermal growth factor receptor/ErbB-2 tyrosine kinase inhibitor, GW2016, on the growth of human normal and tumor-derived cell lines in vitro and in vivo. Mol Cancer Ther. 2001;1:85–94. 25. Escudero-Ortiz V, Pérez-Ruixo JJ, Valenzuela B. Development and validation of a simple and sensitive HPLC-UV method for lapatinib quantification in human plasma. Ther Drug Monit. 2013;35:796–802. 26. Nabors LB, Supko JG, Rosenfeld M, et al. Phase I trial of sorafenib in patients with recurrent or progressive malignant glioma. Neuro Oncol. 2011;13:1324–1330. 27. Navid F, Baker SD, McCarville MB, et al. Phase I and clinical pharmacology study of bevacizumab, sorafenib, and low-dose cyclophosphamide in children and young adults with refractory/recurrent solid tumors. Clin Cancer Res. 2013;19:236–246.
325
