A Validated Chiral LC–MS/MS Method for the Enantioselective Determination of (S)-(+)- and (R)-(-)-Ibuprofen in Dog Plasma: Its Application to a Pharmacokinetic Study

The purpose of this study was to develop a method for simultaneously separating ibuprofen enantiomers using electrospray ionization (ESI) liquid chromatography with tandem mass spectrometry (LC–MS/MS). LC–MS/MS was operated with negative ionization and multiple reaction monitoring modes; transitions were monitored at m/z of 205.1 > 160.9 for ibuprofen enantiomers, 208.1 > 163.9 for (S)-(+)-ibuprofen-d3 [internal standard 1 (IS1)], and 253.1 > 208.9 for (S)-(+)-ketoprofen (IS2), respectively. In a one-step liquid–liquid extraction, 10 μL plasma was extracted with ethyl acetate:methyl tertiary-butyl ether of 7:3. Enantiomer chromatographic separation was carried out with an isocratic mobile phase consisting of 0.008% formic acid in water–methanol (v/v) at a flow rate of 0.4 mL/min on a CHIRALCEL® OJ-3R column (150 × 4.6 mm, 3 µm). This method was fully validated for each enantiomer and results were in compliance with the regulatory guidelines of the U.S. Food and Drug Administration and the Korea Ministry of Food and Drug Safety. The validated assay was executed for nonclinical pharmacokinetic studies after oral and intravenous administration of racemic ibuprofen and dexibuprofen in beagle dogs.


Compounds
Ion

Plasma Sample Preparation
Plasma samples treated with heparin were thawed at room temperature on the day of analysis. Each plasma sample (10 µL) was placed in a microtubule. IS1 (20 µL, 10 µg/mL), IS2 (20 µL, 10 µg/mL), 0.1% formic acid (50 µL), and 1 mL of EtOAc:MTBE = 7:3 (v/v) were added to each microtube. After vortexing for 5 min and centrifuging at 20,800 g for 10 min, the supernatant (1.0 mL) was moved to a clean microtube, evaporated, and dried under N 2 gas at 40 • C. The residues were reconstituted in 100 µL of 50% methanol. For analysis, 10 µL of the supernatant was injected into the analytical column.

Method Validation
The developed bioanalytical method was validated in terms of selectivity, lower limits of quantification, linearity, precision, accuracy, recovery, matrix effect, stability, carryover, and dilution integrity. Bioanalytical assay validation was accomplished according to the method validation guidelines of the U.S. Food and Drug Administration (USFDA) and the Ministry of Food and Drug Safety in Korea (MFDS) [44,45].

Selectivity and Sensitivity
Selectivity is a method for determining whether the analyte and IS are quantifiable separately from the other substances present in the sample. Selectivity was determined by analyzing seven randomly selected blank beagle dog plasma samples. Method selectivity and sensitivity were evaluated by comparing chromatograms of blank plasma, plasma spiked with IS1 (10 µg/mL) or IS2 (10 µg/mL), plasma spiked with (S)-(+)-ibuprofen (80 µg/mL) or (R)-(-)-ibuprofen (80 µg/mL), and plasma spiked with (S)-(+)-ibuprofen (0.1 µg/mL), (R)-(-)-ibuprofen (0.1 µg/mL), IS1(10 µg/mL), and IS2 (10 µg/mL). The lower limit of quantification (LLOQ) was described as the lowest concentration of the calibration curve with a signal-to-noise ratio (S/N) greater than 10. The limit of detection (LOD) was described as the lowest concentration of mass detected with an S/N ratio greater than 3.

Linearity and Carryover
Calibration curves of the standards were composed at seven (S)-(+)-ibuprofen and (R)-(-)-ibuprofen concentrations (0.1, 0.5, 2, 10, 20, 40, and 80 µg/mL). Linearity was determined by the calculated peak area ratios (x) of the standard to IS versus the concentrations of the standard (y) using weighted (1/x 2 ) linear least squares regression (y = ax + b). A calibration curve with a correlation coefficient (r 2 ) of 0.99 or greater was considered adequate. Carryover evaluates whether the residual analyte of the previously injected sample before the analysis affects the injection of the next sample during continuous measurement of the sample. Carryover was evaluated by injecting a blank sample after injecting an upper limit of quantification (ULOQ) of the calibration curve standard sample.

Precision and Accuracy
Inter-and intra-assays were performed using replicate analyses of validation sample concentrations (0.1, 0.3, 30, and 64 µg/mL) on three consecutive days. The mean and standard deviation (SD) of the concentrations calculated for these batches were estimated. Accuracy and precision were calculated as relative error and coefficient of variation (CV), respectively. The acceptance criterion was within ±15% of the nominal concentration, except for the LLOQ, which was within ±20%.

Recovery and Matrix Effect
The recovery and matrix effects were evaluated at three QC concentrations and analyzed using analytes spiked before the pre-extraction matrix ([A]), analytes spiked after the post-extraction matrix ([B]), and pure analyte solutions in 50% methanol ([C]). The recoveries of (S)-(+)-ibuprofen, (R)-(-)-ibuprofen, and ISs at QC concentrations were evaluated using the ratio (A/B × 100)%. The matrix effect of (S)-(+)-ibuprofen, (R)-(-)-ibuprofen, IS1, and IS2 associated with ion suppression or enhancement caused by the plasma matrix was assessed using the ratio (B/C × 100)%.

Stability
The stability of the stock and working solutions was evaluated in triplicate for low and high QC concentrations stored at room temperature for 3 and 7 h, respectively, by comparing their peak areas with those of freshly prepared stock and working solutions. The stability of (S)-(+)-ibuprofen and (R)-(-)-ibuprofen in plasma was evaluated by analyzing samples in triplicate for each QC concentration of 0.3, 30, and 64 µg/mL under the following experimental conditions: freeze-thaw stability after three freeze-thaw cycles at −70 • C; short-term stability at room temperature, 4 • C, and −70 • C for 7 h; auto-sampler stability at 10 • C for 54 h; and long-term stability at −70 • C for 196 days.

Dilution Integrity
Dilution integrity was investigated for samples with quantified concentrations outside the calibration curve range. Beagle dog plasma samples with (S)-(+)-ibuprofen and (R)-(-)ibuprofen concentrations above the ULOQ (80 µg/mL) were diluted with blank beagle dog plasma and reanalyzed. Diluted QC samples of (S)-(+)-ibuprofen and (R)-(-)-ibuprofen, which had 5-fold higher concentrations than those of the QC 0.3, 30, and 64 µg/mL, were prepared as described in Section 2.3. The diluted QC samples were diluted five times with blank beagle dog plasma at the original QC concentration and preparation, as described in Section 2.4. Precision and accuracy were assessed in terms of CV (acceptable range: <15%) and RE (acceptable range: within ±15%), respectively.
Pharmacokinetic parameters including extrapolated AUC inf (area under the plasma concentration-time curve from time zero to infinity), AUC last (area under the plasma concentration-time curve from time zero to the time of the last measurable concentration), C max (maximum plasma drug concentration), T max (the time to reach C max ), terminal halflife (t 1/2 ), apparent total clearance of the drug from plasma after oral administration (CL/F), apparent volume of distribution during the terminal phase after oral administration (V z /F), and mean residence time (MRT) were determined using noncompartmental methods (WinNonlin version 8.1; Pharsight Corporation, Mountain View, CA, USA).

Incurred Sample Reanalysis (ISR)
ISR was performed via computerized selection (sampling without replacement) of 15 subject samples near C max and at the elimination phase of the pharmacokinetic profile. The results were compared to the initial data obtained from the same samples using the same procedure. The percentage change in the measured values did not exceed ± 20%.

Mass Spectrometry
An infusion was performed to optimize mass spectrometric conditions. For infusion, 100 ng/mL (S)-(+)-ibuprofen, (R)-(-)-ibuprofen, IS1, and IS2 solutions were injected into the mass spectrometer using a syringe pump at a flow rate of 10 µL/min. The maximum intensities of the product and fragment ions of ibuprofen and IS were obtained with an electrospray (ESI) negative interface because the negative mode of MS detection conditions was better ionized, more sensitive, and had no interference. Full-scan mass spectra were characterized by deprotonated molecules [M-H] − at m/z 205.1 for (S)-(+)-ibuprofen and (R)-(-)-ibuprofen and m/z 208.1 for IS1 and m/z 253.1 for IS2. The selected product ion m/z were 160.9 for both (S)-(+)-ibuprofen and (R)-(-)-ibuprofen, 163.9 for IS1, and 208.9 for IS2 ( Figure 1). The ionization source settings of the mass parameter according to the signal intensity were optimized by flow injection analysis tuning. The optimized mass parameters of the temperature and ion spray voltage were 450 • C and −4500 V, respectively, bringing about the highest signal intensity among the tested conditions. Chromatography revealed that analytes and IS fragments did not interfere with each other.
profen and (R)-(-)-ibuprofen and m/z 208.1 for IS1 and m/z 253.1 for IS2. The selected product ion m/z were 160.9 for both (S)-(+)-ibuprofen and (R)-(-)-ibuprofen, 163.9 for IS1, and 208.9 for IS2 ( Figure 1). The ionization source settings of the mass parameter according to the signal intensity were optimized by flow injection analysis tuning. The optimized mass parameters of the temperature and ion spray voltage were 450 °C and −4500 V, respectively, bringing about the highest signal intensity among the tested conditions. Chromatography revealed that analytes and IS fragments did not interfere with each other.

Chromatographic Conditions
We achieved an appropriate peak shape, separation, and run time through developing the chromatographic conditions for separating ibuprofen enantiomers, including the columns, column temperature, mobile phase composition, and flow rate. In this study, the following reversed-phase columns and chiral columns were evaluated for favorable separation: DAICEL CHIRALPAK ® AGP (150 × 2.0 mm, 5 μm), DAICEL CHIRALCEL ® OJ-RH (150 × 4.6 mm, 5 μm), DAICEL CHIRALCEL ® IJ-3 (150 × 4.6 mm, 3μm), and DAICEL CHIRALCEL ® OJ-3R (150 × 4.6 mm, 3 μm). Among the tested columns, the DAICEL CHIRALCEL ® OJ-3R column exhibited the best results regarding peak shape, separation, sensitivity, and retention time. Chiral columns, such as α-acid glycoprotein (AGP), OJ-RH, and IJ-3 columns did not show complete separation, leading to inadequate peak shape and sensitivity. Mobile phases containing diverse buffers, such as formic acid and ammonium acetate, were evaluated using gradient and isocratic methods. The optimal column temperature and mobile phase flow rates were determined in various trials regarding peak sensitivity and chromatographic separation of enantiomers. The retention time of the peak was unstable when a gradient mobile phase was used with two different pumps. Therefore, as suggested in previous studies using the LC-MS/MS system [38][39][40][41][42], an isocratic mobile phase was used with the aqueous solvent mixed with the organic

Chromatographic Conditions
We achieved an appropriate peak shape, separation, and run time through developing the chromatographic conditions for separating ibuprofen enantiomers, including the columns, column temperature, mobile phase composition, and flow rate. In this study, the following reversed-phase columns and chiral columns were evaluated for favorable separation: DAICEL CHIRALPAK ® AGP (150 × 2.0 mm, 5 µm), DAICEL CHIRALCEL ® OJ-RH (150 × 4.6 mm, 5 µm), DAICEL CHIRALCEL ® IJ-3 (150 × 4.6 mm, 3µm), and DAICEL CHIRALCEL ® OJ-3R (150 × 4.6 mm, 3 µm). Among the tested columns, the DAICEL CHIRALCEL ® OJ-3R column exhibited the best results regarding peak shape, separation, sensitivity, and retention time. Chiral columns, such as α-acid glycoprotein (AGP), OJ-RH, and IJ-3 columns did not show complete separation, leading to inadequate peak shape and sensitivity. Mobile phases containing diverse buffers, such as formic acid and ammonium acetate, were evaluated using gradient and isocratic methods. The optimal column temperature and mobile phase flow rates were determined in various trials regarding peak sensitivity and chromatographic separation of enantiomers. The retention time of the peak was unstable when a gradient mobile phase was used with two different pumps. Therefore, as suggested in previous studies using the LC-MS/MS system [38][39][40][41][42], an isocratic mobile phase was used with the aqueous solvent mixed with the organic solvent to reduce the variation in the separation performance of the LC system. Among the tested solvents for the mobile phase, acetonitrile resulted in an excessively high baseline and delayed analyte elution. Methanol also exhibited a high baseline, but when formic acid buffer was added to the buffer, the baseline was considerably lowered and stabilized. Therefore, methanol was selected as the optimal organic solvent for the mobile phase, which was determined based on the peak shape and resolution. Depending on the concentration of formic acid, a different sensitivity and chiral separation by the interference of baseline was found. During the development of our method using various concentrations of formic acid, the optimal concentration of formic acid was determined to be 0.008% for achieving both chiral separation and sensitivity. The best separation results were obtained with the OJ-3R column at a temperature of 40 • C using an isocratic mobile phase constituting 85% methanol (water:methanol = 15:85, v/v) with 0.008% formic acid at a flow rate of 0.4 mL/min. Overall, the optimized chromatographic conditions mentioned above achieved the best peak shape, separability, sensitivity, and reproducibility.

Sample Preparation
Various methods, such as solid-phase extraction (SPE), (liquid-liquid extraction) LLE, and derivatization, have been used to extract ibuprofen [38][39][40][41][42][43]. A previous study analyzing ibuprofen enantiomers found lower matrix effects with LLE than with SPE when extracting ibuprofen from biological samples [41]. In addition, LLE is a simpler and faster extraction method than the derivatization or SPE methods; therefore, a simple, one-step LLE extraction was utilized without the derivatization process as the extraction method in our study. Various conditions were tested to obtain better recovery rates and sensitivities using less plasma than previously reported methods [38][39][40][41][42][43]. To determine the optimal extraction conditions with respect to recovery and matrix effects, various extraction solvents were evaluated along with methylene chloride, EtOAc, n-hexane, MTBE, and mixtures of MTBE:EtOAc (8:2, 7:3, 3:7, 2:8; v/v) and EtOAc:n-hexane (8:2, 7:3, 3:7, 2:8; v/v). Extraction buffers, such as acetic acid, formic acid, hydrochloric acid, and reconstitution solvents, including methanol and acetonitrile, were assessed to achieve the optimal analyte peak shape and intensity. The best peak intensity and analyte recovery were obtained using 1 mL of MTBE:EtOAc (3:7; v/v) with 50 µL of 0.1% formic acid buffer as the extraction solvent and 50% methanol as the reconstitution solvent in our present study. Therefore, the developed plasma sample extraction method is a fast and simple extraction procedure with good sensitivity using a small amount of plasma (10 µL); thus, an abundant number of plasma samples can be analyzed.

Application to a Pharmacokinetic Study
Dexibuprofen ((S)-(+)-ibuprofen), an active enantiomer of ibuprofen, is available as an oral tablet in the market [9,10,46]. In this study, the developed and validated LC-MS/MS method was successfully implemented in a pharmacokinetic study to analyze 144 plasma samples from beagle dogs after oral and intravenous administration of racemic ibuprofen and (S)-(+)-ibuprofen. The mean ± SD plasma concentration-time profiles of (S)-(+)-ibuprofen and (R)-(-)-ibuprofen in the beagle dogs are shown in Figure 4. Table 6 summarizes the pharmacokinetic parameters of racemic ibuprofen and (S)-(+)-ibuprofen. Non-compartmental analyses were performed to calculate the following pharmacokinetic parameters using the noncompartmental methods (WinNonlin version 8.1; Pharsight Corporation, Mountain View, CA, USA): extrapolated AUC inf , AUC last , C max , T max , t 1/2 , CL/F, V z /F, and MRT.
Compared to previous studies, the enantiomeric ratio of AUC (S)-(+)-ibuprofen / AUC (R)-(-)-ibuprofen estimated in our current study was smaller (2.09 after oral and 2.55 after intravenous administration vs. 3.27 to 6.00 in previous studies) [29,40,54]. The degree of R-to-S inversion may vary depending on the dosage form [51,54] and study subject species; the rate of isomeric inversion may vary in different species with different mecha-nisms of isomeric inversion, as well as with different routes of administration [55]. Similar to the inference of our current study, Frihmat et al. suggested a lower isomeric inversion ratio after oral administration than after intravenous administration of racemic ibuprofen at the same dose (35.34% vs. 43.62%) in beagle dogs [52]. Overall, these results suggest a lower isomeric inversion rate with oral administration than with intravenous administration [49]. Consistent with previous studies in humans [26,48,50] and dogs [52,53], (R)-(-)-ibuprofen was not detected in the plasma after oral and intravenous administration of (S)-(+)-ibuprofen, suggesting that there was no significant isomeric inversion from (S)-(+)-ibuprofen to (R)-(-)-ibuprofen. Although previous research has suggested bioequivalence between (S)-(+)-and racemic ibuprofen tablets after oral administration of 400 mg to humans [26], no pharmacokinetic or bioequivalence evaluation has been performed for (S)-(+)-ibuprofen after intravenous injections of (S)-(+)-and racemic ibuprofen. Our present study exhibited higher C max , AUC last , and AUC inf of (S)-(+)-ibuprofen after intravenous administration of 300 mg (S)-(+)-ibuprofen than after that of 400 mg racemic ibuprofen, suggesting a potentially higher pharmacological activity of the newly developed (S)-(+)-ibuprofen injection than the racemic ibuprofen injection.

Incurred Sample Reanalysis (ISR)
ISR was executed to determine the reproducibility of the developed analytical method in the present study. Conforming to regulatory guidelines [44,45], the bioanalytical assay was regarded as reproducible if at least 67% of the tested samples had a deviation within ±20% between the original measurement and the reanalysis results. Two of the fifteen ISR samples exceeding the ULOQ were further diluted and reanalyzed as described in Section 3.2.6. All the reanalyzed samples (n = 15) satisfied the pre-described regulatory ISR acceptance criteria [56] for both (S)-(+)-ibuprofen and (R)-(-)-ibuprofen (Table S2).

Conclusions
In this study, a selective and reproducible analytical method was developed and validated to determine the concentration of ibuprofen enantiomers in beagle dog plasma. The developed method was fully validated according to the MFDS and USFDA guidelines [44,45], including the specificity, reliability, and reproducibility of the established quantitation method over the concentration range. In addition, our newly developed method was successfully applied to pharmacokinetic research for the simultaneous quantification of ibuprofen enantiomers after oral and intravenous administration of (S)-(+)-ibuprofen and racemic ibuprofen in beagle dogs (tablets for oral administration, injection solutions for intravenous administration). Based on the observed differences in plasma concentrations of (S)-(+)-ibuprofen following oral or intravenous administration of (S)-(+)-ibuprofen and racemic ibuprofen, the pharmacokinetic profiles of the two enantiomers were substantially different. The unidirectional inversion of R-(-)-ibuprofen to (S)-(+)-ibuprofen after the administration of racemic ibuprofen might contribute to the pharmacokinetic differences between the two enantiomers. Therefore, our findings may assist in understanding the pharmacokinetic properties of ibuprofen after oral and intravenous administration to support further clinical development of (S)-(+)-ibuprofen intravenous injection solutions.