Determination of clonixin in plasma and urine by reversed-phase high-performance liquid chromatography

Journal of Chromatography A, 889 (2000) 135–141 www.elsevier.com/locate/chroma Determination of clonixin in plasma and urine by reversed-phase high-performance liquid chromatography aa,b c,d * ˜ ´´ Antonio Bica , Ascensao Farinha , Henning Blume , Carlos Maurıcio Barbosa a ´ˆ Laboratorio de Estudos Farmaceuticos , Rua Alto do Duque 67, 1400 Lisbon , Portugal b ¨ Johann Wolfgang Goethe-Universitat , Frankfurt am Main , Germany c ´ CTMUP , Faculdade de Farmacia da Universidade do Porto , Porto , Portugal d ´ CETMED , Centro Tecnologico do Medicamento , Porto , Portugal Abstract A reversed-phase high-performance liquid chromatographic method that enables the determination of clonixin in human plasma and urine samples is described. Recovery of the drug was over 87.6 and 80.7% for plasma and urine, respectively. The limit of quantitation of the method was established as 10 ng/ml in plasma and 20 ng/ml in urine samples, with RSDs of less than 11.1%. The applicability of the method was further assessed by determining the plasma concentrations time course of clonixin in six healthy volunteers after single oral dose administration of 150 and 300 mg of clonixin and Clonix. 2000 Elsevier Science B.V. All rights reserved. Keywords : Validation; Clonixin; Nonsteroidal anti-inflammatory drugs 1. Introduction activity in laboratory animals [1]. Analgesic properties have been shown in monkeys [2] and in Clonixin, 2-(3-chloro-o-toluidino) nicotinic acid is humans [3–10]. Clonixin showed a good safety a nonsteroidal anti-inflammatory drug (Fig. 1) disprofile, with analgesic efficacy, following oral adplaying anti-inflammatory, analgesic and antipyretic ministration of a 600-mg dose, comparable to morphine sulphate 10 mg parenterally [9]. The drug’s solubility in aqueous media is very low, with a high trend for partitioning into organic solvents, as revealed by its high octanol/water partition coefficients (unpublished data). Few methods for the assay of clonixin in biological samples have been reported. Katchen and co-workers [11,12] developed a thin-layer chromatographic method following UV spectrophotometry to study the metabolism and pharmacokinetics of Fig. 1. Clonixin chemical structure. clonixin in rats, dogs and monkeys and in humans, following i.v. and oral administration of 23250 mg gelatine capsules. Clonixin was found to be rapidly *Corresponding author. E-mail address : [email protected] (A. Farinha). absorbed and excreted by humans. The excretion 0021-9673/00/$ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0021-9673(00)00348-4 136 A . Bica et al . /J . Chromatogr . A 889 (2000) 135 – 141 patterns of the parent compound and the urine 2.3. Plasma sample preparation metabolites were very similar between rhesus monkeys and humans. A 1-ml plasma sample, thawed at room tempera- ´ Gonzalez-Martin et al. [13], reported a study of ture, was spiked with 50 ml of a 21.4-mg/ml I.S and the pharmacokinetics of the lysine salt of clonixin in mixed with 50 mlof6Mhydrochloric acid in a glass 10 children in postoperative care, aged 4–10 years screw-capped tube. To each sample, 4 ml of a old. Serum clonixin concentrations were determined mixture of chloroform–methanol (99.5:0.5) were by a revered-phase high-pressure liquid chromatoadded and the tubes thoroughly mixed for 15 min. graphic method with a limit of quantitation (LOQ) of After centrifugation at 4500 rpm for 5 min, the 100 ng/ml, although further details were not proorganic layer was transferred to a clean glass tube. vided. The pharmacokinetic parameters determined The remaining aqueous layer was again extracted were different from those previously reported in with additional 4 ml of the chloroform–methanol adult humans and experimental animals. solution. The organic layer was transferred to the We report here a new specific and sensitive tube containing the first 4 ml organic phase from the method for the determination of clonixin in human previous extraction step. The pooled organic layers plasma and urine samples. The method was applied were evaporated to dryness at 458C under a stream of to a pilot bioavailability study of clonixin in six nitrogen. The dry residue was reconstituted with 100 healthy volunteers. ml of methanol. Twenty-ml samples were injected into the HPLC apparatus. 2.4. Urine sample preparation 2. Experimental Frozen urine was allowed to thaw at room temperature. The internal standard (p-chlorowarfarin 2.1. Materials solution at 21.4 mg/ml), 50 mlof6Mhydrochloric acid, 50 mg of sodium chloride and 4 ml of a Clonixin was a generous gift of Janssen-Cilag mixture of chloroform–methanol (99.5:0.5) were (Amadora, Portugal). p-Chlorowarfarin (internal added to a 1.0-ml urine aliquot in a screw-capped standard, I.S.) was supplied by Aldrich (Madrid, glass tube. The tubes were thoroughly mixed for 15 Spain). All solvents were HPLC grade from E. min. After centrifugation at 4500 rpm for 5 min, the Merck (Darmstadt, Germany). Other reagents were organic layer was transferred to a clean glass tube of analytical-reagent grade, also from E. Merck. and evaporated to dryness under a stream of nitrogen at 458C. The residue was reconstituted with 100 ml 2.2. High-performance liquid chromatography of methanol and 20 ml were injected into the HPLC ( HPLC ) apparatus. The HPLC pumping system used in this study was 2.5. Validation study a Merck–Hitachi L-6000A pump. The mobile phase was 28% acetonitrile, 36% methanol and 36% of To assess linearity in plasma samples, two cali0.3% acetic acid solution at a flow-rate of 1.0 ml/ bration curves were prepared. A lower calibration min. Sample injection was performed with a Waters curve ranging from 0.010 to 2.0 mg/ml and a upper 717 Plus autosampler. The HPLC column (25034.0 calibration curve ranging from 0.10 to 26.6 mg/ml. mm I.D.) was a Nucleosil-100, 5-mm C column, A single calibration range was prepared in urine 18 coupled to a mBondapak C guard-column. A from 0.020 to 2.0 mg/ml. Standard samples were 18 Waters 486 UV detector set at 290 nm (0.005 AUFS) prepared by spiking drug-free plasma and urine with was used to monitor the column effluent. A Merck– clonixin and the internal standard following extracHitachi D-2500 Chromato-Integrator was used to tion and analysis as described above. The peak acquire and integrate the chromatographic data. height ratios of clonixin to the internal standard were A . Bica et al . /J . Chromatogr . A 889 (2000) 135 – 141 137 measured and the calibration curves were obtained had the approval of the clinical unit Ethics Commit2 by least-squares weighted (1/conc ) linear regrestee. sion. Regression lines were used to determine clonixin concentrations in the test samples. 2.7. Pharmacokinetic analysis The absolute recovery of clonixin extracted from urine and plasma samples was assessed by comparAll pharmacokinetic parameters were determined ing the peak heights of extracted plasma and urine by non-compartmental models. The slope of the samples, containing known amounts of drug, with terminal log–linear portion of the clonixin concenthose obtained from standard solutions of clonixin, tration versus time profiles was determined by the prepared in methanol–water (1:1), at the same best fit least-squares regression analysis and used as concentration levels. The concentrations of clonixin the elimination rate constant (K). The apparent e tested in plasma samples were 0.010, 0.020, 0.40 and elimination half-life (t) was determined as ln 2/ 1/2 2.0 mg/ml, while in urine samples the concentration K. The area under the concentration versus time e levels were 0.020, 0.40 and 2.0 mg/ml. curve of clonixin in plasma from time zero to the last Intraand inter-assay RSDs and relative errors quantifiable point (AUC ) was calculated by the 02t (REs) were used to assess the precision and accuracy linear trapezoidal rule, from time zero to t, and by max of the methods. In plasma samples these concenthe log-linear trapezoidal rule, beyond tuntil the max tration levels were 0.010, 0.020, 0.40, 2.0 and 8.5 last quantifiable point (C). Extrapolated area under t mg/ml. Additional evaluation for precision and the curve, from Cto infinity (AUC ) was dett2inf accuracy was performed at 15.9 and 26.6 mg/ml. In termined as C/K. Total area under the curve te urine samples the tested levels were those used for (AUC ) was the sum of AUC and AUC . 02inf 02tt–inf assessing absolute recovery. The maximum plasma concentration (C) and the max Validation studies also included specificity tests in time to reach C(t) were taken directly from max max six individual plasma and urine samples, room the observed concentration versus time data. temperature stability, long-term stability and freeze– thaw stability. 3. Results and discussion 2.6. Application of the plasma method The chromatograms of blank plasma, spiked plasThe analytical procedure was applied to the analyma with clonixin and internal standard and plasma sis of plasma samples from a pilot pharmacokinetic from a subject dosed with Clonix, are shown in Fig. and bioavailability study of clonixin, following the 2a–c, respectively. Chromatograms of blank urine administration of Clonix (Janssen-Cilag) 300-mg and urine spiked with clonixin and internal standard immediate release gelatine capsules, from a commerare shown in Fig. 3a,b. No interfering peaks were cially available batch (no. 7G128A; expiry date, July observed at the clonixin retention times. 2000) and two, 150 and 300 mg, in house-prepared The absolute recovery of clonixin from plasma oral aqueous solutions. Six healthy male volunteers, and urine samples (Table 1) ranged from 87.6 to aged 22–31 years, participated in the clinical in96.4% and 80.7 to 90.1%, respectively. All calivestigations. None of the subjects had any relevant bration lines showed to be linear over the conmedical history or was on any other medication. centrations range examined with a r.0.9953. Subjects fasted overnight prior to the administration Tables 2 and 3 include the data obtained for the at 08:00 h. Blood samples were withdrawn until 14 h precision and accuracy. Maximum intraand interpost-administration and transferred to lithium heparin day precision was 11.06 and 8.07% for plasma and Luer Monovette tubes (Sarstedt, Portugal). The urine, respectively. The accuracy observed for plasresulting plasma (2500 rpm for 10 min, 58C) was ma and urine samples, ranged from 21.67 to stored frozen below 2208C in polypropylene tubes, 10.70% and from 21.90 to 7.68%, respectively. pending analysis. All the subjects gave informed Additional data obtained during the analysis of written consent prior to the start of the study, which plasma samples from the pilot study, showed RSDs 138 A . Bica et al . /J . Chromatogr . A 889 (2000) 135 – 141 Fig. 2. Typical chromatograms of a blank plasma sample (a), a plasma spiked with clonixin and I.S. (b) and a sample from a Fig. 3. Typical chromatograms of a blank urine sample (a) and a subject dosed with Clonix (c). urine sample spiked with clonixin and I.S. (b). A . Bica et al . /J . Chromatogr . A 889 (2000) 135 – 141 139 Table 1 ´ times lower than that reported by Gonzales-Martin et Recovery of clonixin from plasma and urine al. [13], for serum samples. Concentration Recovery (%) RSD (%) Clonixin proved to be stable under different added (mg/ml) conditions. After three freeze–thaw stability cycles Plasma 1.996 87.6 14.6 of samples spiked with 0.4 mg/ml of clonixin (n5 (n56) 0.399 88.5 6.3 6), a negligible difference was observed from time 0.020 96.4 7.1 zero to the last cycle. At room temperature, clonixin 0.010 91.3 5.5 showed to be stable for at least 24 h, both in plasma Urine 2.005 88.0 5.3 and urine samples at 0.4 mg/ml (n56), with 23.3 (n56) 0.400 90.1 4.3 and 24.6% differences observed from time zero to 0.020 80.7 12.9 24 h after sample preparation. Long-term stability of stored samples at or below 2208C was assessed by analysing six replicate of 3.22% (n56) and 6.79% (n58), at 16.0 and 26.6 stability samples at a nominal concentration of 0.4 mg/ml, respectively, while accuracy was 1.29% and mg/ml in comparison with daily prepared samples at 20.06%, respectively. The LOQ was defined as the the same concentration level. Plasma samples proved minimum determinable concentration with a relative to be stable for at least 14 weeks (23.4%), while error below 20% [14]. This was 10 ng/ml (RE, urine samples were stable for at least 4 weeks 21.67%) and 20 ng/ml (RE, 21.90%), respectively, (13.5%). These results indicate a good stability of in plasma and urine. This limit of quantification is 10 clonixin in both types of biological matrix. So far, no Table 2 Within and between-day precision and accuracy for plasma samples Concentration Concentration found RSD (%) RE (%) added (mg/ml) (mg/ml) Within-day 8.506 9.41660.6343 6.74 10.70 (n56) 1.996 2.07160.0390 1.88 3.74 0.399 0.42560.0138 3.26 6.40 0.020 0.019860.0010 5.18 21.25 0.010 0.009860.0011 11.06 21.67 Between-day 8.506 8.61960.5351 6.21 1.33 (n512) 1.996 2.08660.1441 6.91 4.52 0.399 0.40360.0379 9.40 1.11 0.020 0.019760.0013 6.47 21.63 0.010 0.009860.0008 8.32 21.58 Table 3 Within and between-day precision and accuracy for urine samples Concentration Concentration found RSD (%) RE (%) added (mg/ml) (mg/ml) Within-day 2.005 2.02160.0218 1.08 0.77 (n56) 0.400 0.41460.0203 4.89 3.48 0.020 0.019660.0015 7.72 21.90 Between-day 2.005 2.11860.0933 4.40 5.64 (n518) 0.400 0.43160.0237 5.50 7.68 0.020 0.020260.0016 8.07 1.15 140 A . Bica et al . /J . Chromatogr . A 889 (2000) 135 – 141 Fig. 4. Mean plasma levels of clonixin with AUC and Cresults (n56). max information regarding clonixin stability in biological AUC and C. Similar results were obtained from max samples has been published. the analysis of both oral solutions, with ANOVA log Fig. 4 shows the mean plasma profiles of clonixin intra-subject RSDs of 15.5 and 18.8%, respectively, (n56) following administration of the three study for AUC and C. 02inf max formulations. Geometric mean AUC values (n5The results reported clearly show that the val02inf 6) represent essentially true experimental values, as idated HPLC method for plasma samples is applicthe mean extrapolated AUC , was 0.01% for both able for the determination of clonixin in human t2inf Clonix and the 300 mg oral solution and 0.03% for plasma samples from pharmacokinetic studies. the 150 mg oral solution. As indicated by the C max values, clonixin absorption was considerably retarded after administration of the solid oral formulaReferences tion, when compared with the oral solutions. However, the rate of absorption remained unchanged [1] A.S. Watnick, R.I. Taber, I.I. Tabachnick, Arch. Int. Pharwhen the same formulation was given at two differmacodyn. 190 (1971) 78. ent dose levels. [2] V.B. Ciofalo, J. Patel, R.I. Taber, Jpn. J. Pharmacol. 22 21 (1972) 749. The plasma elimination rate constant (0.45 h ) [3] H. Frerick, Zahnarztliche Praxis 10 (1994) 342. was very similar for all three formulations and in ´ [4] M.L. Marti, C. Fraga, J.A. Pezza, Prensa Med. Argentina S8 accordance with that previously reported [11], fol- (1993) 695. lowing single oral dose administration of a solid [5] G. Nassoweter, Presented at the 1st International Conference preparation. The apparent elimination half-lives were on Pain, Buenos Aires, 19–20 April. [6] M. Osswald, H. Frerick, J. Maisenbacher, N. Schenk, P. 1.6, 1.7 and 1.5 h, respectively, for Clonix, the Grabner, ZWR 103 (1994) 381. 300-mg and the 150-mg oral solutions. [7] H. Paredes, M. Camus, G. Ballesteros, C. Rondon, Presented Analysis of variance (ANOVA ) intra-subject log at the 7th World Congress on Pain, Paris, 22–27 August. RSDs for the comparison of Clonix versus. the 300 [8] K. Schaffer, C.H. Wauschkuhn, Presented at the 7th World mg solution, were 14.1 and 24.8%, respectively, for Congress on Pain, Paris, 22–27 August. A . Bica et al . /J . Chromatogr . A 889 (2000) 135 – 141 141 [9] J.S. Finch, T.J. DeKornefeld, J. Clin. Pharmacol. 15 (1971) [12] B. Katchen, S. Buxbaum, J. Ning, J. Pharmacol. Exp. Ther. 371. 187 (1973) 152. ˜ ´ [10] J.M.G. Santos Bessa, J.M. Silva e Costa, Rev. Portuguesa [13] G. Gonzalez-Martin, C. Cattan, S. Zuniga, Int. J. Clin. ´ Clın. Ter. 9 (1985) 1. Pharmacol. Ther. 34 (1996) 396. [11] B. Katchen, S. Buxbaum, J. Meyer, J. Ning, J. Pharamacol. [14] V.P. Shah et al., J. Pharm. Sci. 81 (1992) 309. Exp. Ther. 184 (1973) 453.