Determination of ibuprofen in human plasma by high-performance liquid chromatography: validation and application in pharmacokinetic study

BIOMEDICAL CHROMATOGRAPHY
Biomed. Chromatogr. 14: 219–226 (2000)
Determination of Ibuprofen in human plasma by high- performance liquid chromatography: validation and R. Canaparo1*, E. Muntoni1, G. P. Zara1, C. Della Pepa1, E. Berno1, M. Costa2 and M. Eandi11Department of Anatomy, Pharmacology and Forensic Medicine, University of Torino, Torino, Italy 2Department of Electronics, Polythecnic of Torino, Torino, Italy Received 18 February 1999; revised 19 July 1999; accepted 4 August 1999 ABSTRACT: A specific method for the simultaneous determination of S-(‡)Ibuprofen and R-(À)Ibuprofen enantiomers in humanplasma is described. Adopting a high-performance liquid chromatographic (HPLC) system with spectrofluorometer detector, thecompounds were extracted from plasma in alcohol medium and were separated on C18 column, using a solution of acetonitrile–water–acetic acid–triethylamine as mobile phase. The limit of quantitation was 0.1 mg/mL for both compounds. The method wasvalidated by intra-day assays at three concentration levels and was used in a kinetic study in healthy volunteers. During the study wecarried out inter-day assays to confirm the feasibility of the method. Copyright # 2000 John Wiley & Sons, Ltd.
pharmacologically active one, S-(‡) (Adams et al.,1976; Caldwell et al., 1988; Evans, 1992; Hutt and The 2-arylproprionic acid derivative, Ibuprofen [RS-2- Caldwell, 1983; Kaiser et al., 1976; Wechter et al., 1974).
(4-isobutyl-phenyl)propionic acid], is one of the most Up to about 70% of a single oral dose of R-(À)Ibuprofen potent orally active antipyretic, analgesic and non- has been observed to be unidirectionally inverted to its steroidal anti-inflammatory drug (NSAID) used exten- antipode S-(‡)Ibuprofen. The process appears to be very sively in the treatment of acute and chronic pain, rapid in man (Avgerinos et al., 1987; Lee et al., 1985).
osteoarthritis, rheumatoid arthritis and related conditions.
Whether the bioinversion of R-(À)Ibuprofen occurs This compound is characterized by a better tolerability presystemically or systemically remains to be determined compared with other NSAIDs (Busson 1986).
(Cox 1988; Hall et al., 1993; Jamali et al., 1992).
Ibuprofen contains a chiral carbon atom on the Although R-(À)Ibuprofen is inactive, it can interfere with propionic acid side-chain, therefore it exists as two the pharmacokinetic profile of other drugs, as well as enantiomers. It is usually marketed as a 50:50 mixture of with the pharmacokinetics of the S-(‡)enatiomer (Aarons the S-(‡) and R-(À) enantiomers, even if it is known that et al., 1983; Jones et al., 1986; Lee et al., 1985; Lin et al., the pharmacological activity is due almost exclusively to the S-(‡) enantiomer (eutomer) (Adams et al., 1976; The enantiomers of Ibuprofen may compete for protein Caldwell et al., 1988; Evans, 1992; Hutt and Caldwell, binding sites (Jones et al., 1986; Lee et al., 1985; Williams and Le, 1985). Data from non-stereoselective S-(‡)Ibuprofen is more potent than the R-(À)-isomer plasma protein binding studies have shown that Ibupro- (distomer) in inhibiting prostaglandin synthesis with an fen is about 98–99% bound at therapeutic concentrations eudismic ratio (the ratio of activities of the eutomer (Aarons et al., 1983). There is not a great difference versus the distomer) of 160 (Caldwell et al., 1988).
between the enantiomers in the bound fraction; however, Ibuprofen undergoes an in vivo metabolic chiral even a small difference in the bound fraction between the inversion from the inactive R-(À)enantiomer to the enantiomers may represent a substantial difference intheir respective unbound fractions. In plasma the fractionunbound is an important determinant of the primarypharmacokinetic properties of Ibuprofen and of its *Correspondence to: R. Canaparo, Department of Anatomy, Pharma-cology and Forensic Medicine, University of Torino, Via P. Giuria 13, pharmacological activity, since it is generally accepted 10125 Torino, Italy; e-mail: canaparo@medfarm.unito.it that the drug effect is related more closely to the unbound Abbreviations used: ECF, ethyl chlorofomate; EOA, ethanolamine;
fraction than to the total plasma concentrations (Lin et IS, internal standard; LOQ, limit of quantitation; NSAD, non-steroidal al., 1987). Therefore, the bioinversion of R-( anti-inflammatory drug; S-NEA, S( is important for the pharmacokinetic, the therapeutic and/ Copyright  2000 John Wiley & Sons, Ltd.
or the toxicological effects (Caldwell et al., 1988; Hutt Table 1. Pre-study validation in Ibuprofen (R) and (S)
assay. Recovery percentage (n = 10 for interval standard,
In plasma and in other biological fluids several methods for Ibuprofen detection has been developed using high-performance liquid chromatography (Mehvar et al., 1988) with common reverse-phase column. The chiral columns are very expensive, less flexible and less durable (Knihinicki et al. 1990). A few authors suggest derivatization of the samples to study the pharmacoki- netics of the two enantiomers of Ibuprofen with a reverse- phase column (Lemko et al., 1993). However, in our Acceptance criteria (Bressolle et al., 1996; Shah et al., 1992): hands several problems of interferences from the matrix emerged. The aim of this paper is to report an improvedchromatographic method and the validation of the qualityof the results. Therefore we have proposed a new 5 min; the system was allowed to equilibrate with the startingmobile phase for 5 min prior to the next sample injection. The chromatographic method for kinetic study in humans.
flow-rate delivered by the Jasco 980-PU binary pumps was 1.6 mL/min and the eluate was monitored at 280 nm as excitationwavelength and at 320 nm as emission wavelength by a Jasco 821-FP spectrofluorometer detector. Injection of the samples(20 mL) was performed using an autosampler (Jasco 851-AS).
Under these conditions the drugs were eluted at about 9.50 min for Fenoprofen R (internal standard), 14.40 min for Ibuprofenenantiomer S(‡) and 16.50 min for Ibuprofen enantiomer R(À).
Racemic Ibuprofen (R/S) (batch 480/8600) was obtained as giftfrom the Zambon Groups and racemic Fenoprofen (R/S) (batch73H0305) was supplied by Sigma-Aldrich-Flucka, Milan (Italy).
Reagents and solvents, all of analytical grade, were purchased fromCarlo Erba, Turin (Italy) and Aldrich, Milan (Italy). Acetonitrile– The preparation of the samples was performed according to Lemko water–acetic acid–triethylamine (60:40:0.1:0.02 by vol) was used et al. (1993) with minor modification. To 0.5 mL of plasma as mobile phase. The solution of triethylamine 50 mmol (TEA) samples was added 5 mg/mL of internal standard and 200 mL of 1 M was prepared in acetonitrile and the acetic acid was glacial. The sulphuric acid. After brief vortex, 3 mL of isooctane and isopropyl final pH of the mobile phase was 5.0. The solutions of 6 mmol alcohol solution (95:5 v/v) was added and, after brief vortex- ethyl chlorofomate (ECF) and ethanolamine (EOA) 1:40 v/v were mixing, it was centrifuged for 10 min at 1800 g. The organic phase
prepared in acetonitrile. The solution of S-(À)1-(1-naphthyl) was transferred to a clean tube and was evaporated to dryness ethylamine (S-NEA), 0.5 mL/L, was prepared in acetonitrile– under a nitrogen stream in a water bath at 40°C. The residue was TEA (8:2, v/v). These three final solutions were critical for sample redissolved in 300 mL of TEA and after brief, vortex-mixing, 50 mL of ECF was added. After brief vortex-mixing, we added about25 mL of S-NEA. The mixture was vortex-mixed, left standing for 3 min, and 25 mL of EOA was added to stop the ECF reactingfurther. The solution was transferred into an amber glass vial for The drugs were dissolved in acetonitrile and sonicated for automatic injection into the HPLC system.
dissolution, to yield a stock solution of 0.2 mg/mL each. We usedindependently prepared stock solutions for the working racemicIbuprofen, which was dissolved in acetonitrile at concentrations ranging from 50 to 0.1 mg/mL by serial dilution of their respectivestock solutions. Racemic Fenoprofen preparation, ie 5 mg/mL, was We used blank plasma from untreated healthy human blood donors similarly obtained. All solutions were prepared monthly and stored (AVIS, Turin, Italy) spiked with the drug for the construction of refrigerated in the dark when not in use.
calibration curve. Evaluation of the assay was performed with aseven-point calibration curve in the concentration range from 0.1to 50 mg/mL for racemic Ibuprofen. The slope and the intercept of the calibration graphs were calculated through least squares linear regression of each drug to internal standard peak–area ratios vsdrug concentration. Experimental peak–area ratios were inter- Ibuprofen R/S and internal standard Fenoprofen R/S were separated polated on the relative calibration curve and the concentrations on a 4.6 Â 150 mm Symmetry2 C18 (5 mm) column with a Sentry2 guard column (4.6 Â 20 mm, 5 mm) Water, Milan (Italy),operating at room temperature. Elution was performed undergradient conditions. The mobile phase (pH 5), consisting of acetonitrile–water–acetic acid–triethylamine, 60:40:0.1:0.02 (byvol), after a run for 20 min was changed to 100% acetonitrile for The validation of the analytical method was performed according Copyright  2000 John Wiley & Sons, Ltd.
Biomed. Chromatogr. 14: 219–226 (2000)
Table 2. Pre-study validation in Ibuprofen (R) and (S)
Table 4. Pre-study validation in Ibuprofen (R) and (S)
assay. Statistical values. Limit of quantitation (LOQ) 0.1 mg/
assay. Statistical values. Linearity: plasma calibration
mL: plasma spiked samples (n = 5)
curves, descriptive statistics (n = 6) range 0.1–50 mg/m
Acceptance criteria (Bressolle et al., 1996; Shah et al., 1992): accuracy80–120%; precision 20%.
samples stored at À28°C for 3 months, and the Ibuprofen (S) and(R) samples that underwent three cycles of freeze and thaw. The to the suggestions proposed by Shah et al. (1992) and Bressolle et data were calculated using the peak–area ratio Ibuprofen (S)/ internal standard and the Ibuprofen (R)/internal standard. Thesamples stored at À28°C for 3 months were stable and after threecycles of freezing–thawing no degradation was observed (Table 3).
Pre-study evaluationExtraction recovery. Recovery of Ibuprofen (S) and (R) from plasma was measured through the peak–area ratio measurement ofthe compounds in extracted samples and in authentic unextracted Linearity. Intercept, slope and coefficient of correlation (r) were standards, prepared in mobile phase and spiked at two concentra- evaluated for six calibration curves of six independent source of tion levels (1 and 10 mg/mL). The percentage ratio of their peak– plasma and for each calibration curve performed daily. The area (extracted vs unextracted) was taken as the value of extraction calibration was accepted if the (r) value found was above the recovery. Average values were calculated at each concentration tabulated one corresponding to the significant level p 4 0.05 for and at all concentrations tested (Table 1).
the n calibration points and n-2 degrees of freedom (Tables 4 and5).
Limit of quantitation (LOQ, plasma spiked samples). Weperformed replicate analyses (n = 5) of Ibuprofen (S) and (R) in Accuracy and precision (plasma spiked samples). An Intra- plasma samples, spiked at 0.1 mg/mL. The mean coefficient of day assay at concentrations higher than the Limit of Quantitation variation percentage was assumed to be a precision value. The (LOQ) was performed on freshly prepared plasma samples, spiked results were accepted when the precision values were 20% at 0.5, 2 and 15 mg/mL (n b 2/concentration). Acceptance criteria of the results were based on accuracy values, the percentage ratiobetween the mean concentration obtained and the nominal value, in Stability (plasma spiked samples). The stability of Ibuprofen the range from 85 to 115%. Precision values were `20% for low (S) and (R) in plasma samples was tested in storage conditions values and `15% for medium and high values.
(À28°C) for 3 months and after three freeze–thaw cycles. Thestability of Ibuprofen (S) and (R) was evaluated by comparing a Accuracy and precision (plasma calibration samples). The fresh human blank plasma fortified with 1 and 10 mg/mL of concentration value of each calibration point was back-calculated Ibuprofen (R) and (S), and the plasma that underwent three cycles from the equation of the corresponding calibration curve, of freeze and thaw. The stability was calculated as percentage performed daily with the unknown samples. The results were coefficient of variation (CV%) between the Ibuprofen (S) and (R) accepted/rejected according to the evaluation criteria described Table 3. Pre-study validation in Ibuprofen (R) and (S) assay. Stability to different cycles of freezing–thawing
Acceptance criteria (Bressolle et al., 1996; Shah et al., 1992): precision 15%.
Copyright  2000 John Wiley & Sons, Ltd.
Biomed. Chromatogr. 14: 219–226 (2000)
Table 5. Inter-run accuracy and precision (plasma calibration samples)
Back-calculated Ibuprofen (S) plasma concentration (mg/mL):Mean Back-calculated Ibuprofen (S) plasma concentration (mg/mL):Mean Acceptance criteria (Bressolle et al., 1996; Shah et al., 1992) at the lowest concentration: accuracy, 80–120%; precision, 20%; at the otherconcentrations: accuracy, 85–115%; precision 15%.
above. The mean values obtained were statistically evaluated as an plasma samples, the separation of racemic Ibuprofen on a inter-day assay for plasma calibration samples.
C18 column and the wavelength selected were consid-ered preliminary to further improvements of the method.
Accuracy and precision (quality control samples). Before However, in the practical applications, the compounds starting the analysis of the unknown samples, separate aliquots of were seen to overlap with endogenous peaks of the matrix blank plasma samples, spiked at three concentration levels, namely when an isocratic elution was used. To avoid the problem 0.5, 2 and 15 mg/mL of Ibuprofen (S) and (R) were prepared and we employed a solvent select valve which enabled the stored frozen. Two replicates/concentration were thawed daily and late-eluting peak to be rapidly flushed out after every analysed with a complete calibration curve along with unknownsamples. The analysis of the unknown samples was accepted if at sample analysis using 100% acetonitrile, which was least four of the six quality control (QC) samples were found to be delivered at a flow rate of 1.6 mL/min for 5 min. This within Æ 15% for medium and high concentrations and Æ20% for necessitated re-equilibration of the system with the low concentrations, and if the two possible QC outside Æ20% mobile phase for 5 min prior to subsequent analysis.
or Æ15% of their nominal values were not both at the same Nevertheless in the course of the analyses of unknown nominal concentration. The mean, standard deviation and CV%values obtained were considered as an inter-day assay for plasmaQC, and we applied the same statistical criteria of evaluation as inplasma spiked and calibration samples (Table 6).
Table 6. Within-study validation in Ibuprofen (R) and (S)
assay. Statistical values. Inter-run accuracy and precision
(plasma quality control samples)
Even though Ibuprofen is not a new drug, robust analytical methods to be used for pharmacokinetics Ibuprofen (S) concentration found (mg/mL) studies in human volunteers are scarce. The aim of the present work was to develope a new analytical method to be used to study the pharmacokinetic of the two Ibuprofen enantiomers because the enantiomer S(‡) is active and the enantiomer R(À) is inactive, and a part ofthis enantiomer is converted to active form in vivo. The Ibuprofen (R) concentration found (mg/mL) first problem encountered in the development of the analytical method was to separate the two enantiomers from the racemic compound with a C18 column. A further problem was to optimize the concentration and the quantity of the derivatizating agent (S-NEA). The development of our method is based on the work of Acceptance criteria (Bressolle et al., 1996; Shah et al., 1992): Lemko et al. (1993). The extraction procedure of the accuracy, 85–115%; precision, 15%.
Copyright  2000 John Wiley & Sons, Ltd.
Biomed. Chromatogr. 14: 219–226 (2000)
Figure 1. Chromatographic traces obtained from analysis of
Figure 2. Chromatographic traces obtained from analysis of
authentic plasma samples taken from a subject before treatment authentic plasma samples taken from a human donor without (time 0). Conditions: column, Symmetry2 C18 (4.6 Â 150 mm, racemic Ibuprofen and Fenoprofen S (Internal Standard).
5 mm) containing a Sentry2 4.6 Â 20 mm, 5 mm guard column Conditions as described in Fig. 1. Peak identification: plasma detector: spectrofluorometer at 280 nm excitation wavelength without peaks; (I) Fenoprofen S; (II) Ibuprofen (S); (III) and at 320 nm emission wavelength. Peak identification: (I) Fenoprofen (S) (internal standard, 5 mg/mL); (II) byproducts.
plasma samples some subjects (Fig. 1) presented reason we performed some tests to identify the optimal chromatographic profiles quite different from those quantity of S-NEA to be added into the samples. We observed for plasma batches of human donors (Fig. 2), found that 20 mL of derivatizating agent was the optimal showing unexpected interfering peaks in the area of quantity; the reaction with racemic Ibuprofen was good elution of enantiomers of Ibuprofen. This was the main (Fig. 3) and there were no byproducts in the area of reason why we decided to optimize the quantity of the derivatizating agent to be added into the samples. In With these changes, the intra-day analyses carried out previous work Lemko et al. (1993) and Mehvar et al. in pre-study validation confirmed the reproducibility of (1988) suggested adding 25 mL of the derivatizating the method; the data obtained satisfied pre-defined agent at concentrations of 6 mmol into the samples.
acceptance criteria for accuracy and precision. The Nevertheless in some blank plasma we found unexpected method was applied to the analysis of authentic plasma peaks, presumably byproducts of the reaction. For this samples taken from volunteers in a kinetic study during Copyright  2000 John Wiley & Sons, Ltd.
Biomed. Chromatogr. 14: 219–226 (2000)
Figure 3. Chromatographic traces obtained from analysis of
Figure 4. Chromatographic traces obtained from analysis of
authentic plasma sample taken from a subject 7 h after a single authentic plasma sample taken from a subject before treatment oral administration of racemic Ibuprofen, 400 mg tablet.
without racemic Ibuprofen and Fenoprofen (S) (internal Conditions as described in Fig. 1. Peak identification: (I) standard). Conditions as described in Fig. 1. Peak identifica- Fenoprofen (S) (internal standard, 5 mg/mL); (II) Ibuprofen (S); tion: plasma without peaks; (I) Fenoprofen (S); (II) Ibuprofen (S); (III) Ibuprofen (R).
which more than 600 samples were analysed and a back-calculated concentrations of the calibration samples complete within-study assay performed. A good linearity resulted, on average, within the acceptance criteria, was found over the entire range of calibration curves confirming the limits of quantitation previously observed [0.1–50 mg/mL Ibuprofen (R) and Ibuprofen (S); n = 7], [0.1 mg/mL of Ibuprofen (S) and (R), Table 2]. No runs their coefficients of correlation (r) ranging from 0.9992 were rejected, the results obtained in analysis of the for Ibuprofen (S) and 0.9993 for Ibuprofen (R). All the quality control samples, injected daily with unknown runs were acceptable, the values of coefficient of samples, complying with the assumed acceptance criteria correlation being above the significance level p 4 0.05 for seven calibration points and two degrees of freedom.
The applicability of this method was evaluated in the The slope values remained quite similar to those obtained analysis of unknown samples taken from volunteers in in linearity pre-study evaluation testing, thus indicating the kinetic study. We investigated the pharmacokinetics that the method was reproducible (Tables 4 and 5). The of its individual enantiomers, giving 400 mg of racemic Copyright  2000 John Wiley & Sons, Ltd.
Biomed. Chromatogr. 14: 219–226 (2000)
Figure 5. Mean and standard deviation of plasma concentration–
time profiles of Ibuprofen (S) in all healthy subjects after a single
400 mg oral dose of racemic Ibuprofen.
Figure 6. Mean and standard deviation of plasma concentration–
time profile of Ibuprofen (R) in all healthy subjects after a single
400 mg oral dose of racemic Ibuprofen.
Ibuprofen orally to 18 healthy male volunteers. The mean The absorption of both enantiomers was, in general, (ÆSD) plasma concentration–time profiles of total rapid, and the peak concentrations of Ibuprofen (R) and Ibuprofen (S), after the administration of 400 mg of (S) were achieved within 30.93 and 34.18 mg/mL, racemic Ibuprofen, in all volunteers, are presented in Fig.
respectively. The tmax of the two enantiomers R and S 5. The corresponding profiles for total Ibuprofen (R) are were 1.27 and 1.31 hours, respectively. With one ex- ception, the Cmax of Ibuprofen (S) exceeded that Copyright  2000 John Wiley & Sons, Ltd.
Biomed. Chromatogr. 14: 219–226 (2000)
Ibuprofen (R), confirming the metabolic inversion from Caldwell, J., Hutt, A. J. and Fournel Gigleux, S. 1988. Biochemical Ibuprofen (S) to Ibuprofen (R).
Pharmacology 37:105.
Cox, S. R. 1988. Clinical Pharmacological Therapy 43:146.
Evans, A. M. 1992. European Journal of Clinical Pharmcology
Hall, S. D., Rudy, A. C., Knight, P. M. and Braer, D. C. 1993. Clinical Pharmacological Therapy 53:393.
Hutt, A. J. and Caldwell, J. 1983. Journal of Pharmacy and We have described an improved analytical method for the Pharmacology 35:693.
determination of S-(‡) and R-(À)Ibuprofen enantiomers.
Hutt, A. J., and Caldwell, J. 1984. Clinical Pharmacokinetics 9:371.
Jamali, F., Mehvar, R., Russel, A. J., Satari, S., Yakimets, W. W. and
We have also presented a validation procedure of the Koo, J. 1992. Journal of Pharmaceutical Sciences 8:221.
quality of the results. The method was demonstrated to be Jones, M. E., Sallustio, B. C., Purdie, Y. T. and Meffin, P. J. 1986.
highly feasible and reproducible. The applicability of this Journal of Pharmacology and Experimental Therapy 238:288.
Kaiser, D. G., Vangiessen, G. J., Reischer, R. J. and Wechter, W. J.
method was evaluated in the analysis of unknown 1976. Journal of Pharmaceutical Sciences 65:269.
samples taken from volunteers in a pharmacokinetic Knihinicki, R. D., Day, R. O., Graham, G. G. and Williams, K. M.
1990. Chirality 2(3): 134.
Lee, E. J. D., Williams, K., Day, R., Graham, G. and Champion, D.
1985. British Journal of Clinical Pharmacology 19:663.
Lemko, C. H., Caille, G. and Foster, R. T. 1993. Journal of Chromatography 619:330.
Lin, J. H., Cocchetto, D. M. and Duggan, D. E. 1987. Clinical Pharmacokinetics 12:402.
Aarons, L., Grennan, D. M. and Siddiqui, M. 1983. European Journal Mehvar, R., Jamali, F. and Pasutto, F. M. 1988. Clinical Chemistry of Clinical Pharmacology 25:815.
Adams, S. S., Bresloff, P. and Mason, C. G. 1976. Journal of Shah, V. P., Midha, K. K. and Dighe, S. 1992. Journal of Pharmacy and Pharmacology 28:256.
Pharmaceutical Sciences 81:309.
Avgerinos, A. and Hutt, A. J. 1987. Journal of Chromatography Wechter, W. J., Loughead, D. G., Reischer, R. J., Van-Giessen, G. J.
and Kaiser, D. G. 1974. Biochemistry and Biophysics Research Bressolle, F., Bromet-Petit, M. and Audran, M. 1996. Journal of Communications 61:833.
Choromatography B 686:3.
Williams, K. and Le, E. 1985. Drugs 30:333.
Busson, M. 1986. Journal of International Medical Research 14:53.
Copyright  2000 John Wiley & Sons, Ltd.
Biomed. Chromatogr. 14: 219–226 (2000)

Source: http://www.neuronica.polito.it/Data/Sites/1/neuropdf/Biomed/BiomedChrom00_Ibuprofen.pdf