simultaneous determination of leflunomide and its active metabolite, a77 1726, in human plasma by...

��

�� !� �� "�� #�� $�� %��&'

�� !�� "#��$"%&' �� (��

Received 3 November 2002; accepted 4 December 2002

ABSTRACT: The isoxazol derivative leflunomide [N-(4�-trifluoromethylphenyl)-5-methylisoxazole-4-carboxamide] is an inhibitorof de novo pyrimidine synthesis used for the treatment of rheumatoid artrithis. In the present study, a liquid–liquid extraction-basedreversed-phase HPLC method with UV detection was validated and applied for the analysis of leflunomide and its active metabolite,A77 1726, in human plasma. The analytes were separated using a mobile phase, consisting of acetonitrile, water and formic acid (40/59.8/0.2, v/v), at a flow rate of 0.5 mL/min, and UV detection at 261 nm. The retention times for A77 1726, leflunomide and warfarin(internal standard) were 8.2, 16.2 and 12.2 min, respectively. The validated quantification range of the method was 0.05–100 �g/mLfor leflunomide and 0.1–100 �g/mL for A77 1726. The developed procedure was applied to assess steady-state plasma concentrationsof A77 1726 in patients with rheumatoid arthritis treated with 10 or 20 mg leflunomide per day. Copyright � 2003 John Wiley &Sons, Ltd.

KEYWORDS: Isoxazol derivative; metabolism; pharmacokinetics; rheumatoid arthritis

()*+,-./*(,)

Leflunomide [N-(4�-trifluoromethylphenyl)-5-methyl-isoxazol-4-carboxamide] is an isoxazol derivative withboth anti-inflammatory and immunosuppressive proper-ties (Bartlett et al., 1991, 1994; Chong et al., 1993; Silvaet al., 1997; Breedveld and Dayer, 2000). Over the pastfew years, leflunomide has been used as a disease-modifying antirheumatic drug to treat patients withrheumatoid arthritis. The major leflunomide metabolite,A77 1726, has been demonstrated to confer theimmunosuppressive activity of its parent compoundleflunomide (Davis et al., 1996; Silva et al., 1997). Themode of action of A77 1726 is thought to involvereversible inhibition of the enzyme dihydroorotatedehydrogenase, resulting in reduced pyrimidine ribonu-cleotide levels and a decreased proliferation of activatedlymphocytes (Cherwinski et al., 1995; Xu et al., 1997).Apart from this action, leflunomide has been implicatedin inhibition of tyrosine kinase activity of lymphocytes,thereby preventing the activation of T cells (Mattar et al.,1993).

Following oral administration in man, leflunomideundergoes rapid conversion into its active metabolite,A77 1726, via hydrolysis of the isoxazol ring (Green etal., 1995; Davis et al., 1996; Fig. 1). However, thespecific site of leflunomide hydrolysis as well as theparticipating enzymes are unknown as yet, although bothgut wall and liver have been associated with this

Figure 1. Chemical structures of leflunomide and its majormetabolite A77 1726.

BIOMEDICAL CHROMATOGRAPHYBiomed. Chromatogr. 17: 276–281 (2003)Published online in Wiley InterScience (www.interscience.wiley.com).DOI: 10.1002/bmc.244

*Correspondence to: B. Hinz, Department of Experimental andClinical Pharmacology and Toxicology, Friedrich Alexander Uni-versity Erlangen-Nurnberg, Fahrstrasse 17, D-91054 Erlangen,Germany.E-mail: [email protected]

Abbreviations used: CV, coefficient of variation; HPLC, high-performance liquid chromatography; LLOQ, lower limit of quantifica-tion.

Copyright 2003 John Wiley & Sons, Ltd.

ORIGINAL RESEARCH

bioactivating step (Fachinformation Arava�-Filmtablet-ten, Aventis Pharma, Frankfurt/Main, Germany, 2002).To clarify the pharmacokinetics and metabolism ofleflunomide and A77 1726, a sensitive and selectiveanalytical method is needed for the simultaneousdetermination of the parent compound and its metabolite.

In recent years several HPLC methods have beenpublished for the single determination of A77 1726 (Diaset al., 1995; Lucien et al., 1995; Beaman et al., 2002).However, none of these approaches allow the determina-tion of A77 1726 concomitantly with its parent com-pound, leflunomide. This paper describes a simple andsensitive HPLC method with UV detection for thesimultaneous determination of leflunomide and A771726 in human plasma. The developed and validatedHPLC method was applied to assess the steady-statelevels of A77 1726 in plasma of patients with rheumatoidarthritis treated with leflunomide.

!�*0+(�1� �)- !0*%,-�

/�� 2 Acetonitrile (gradient grade), ethylacetate (for liquid chromatography), formic acid 98–100%, sodiumacetate trihydrate, sodium chloride, sodium hydroxide, acetic acidand absolute ethanol (extra pure) were purchased from Merck(Darmstadt, Germany). Warfarin was obtained from Fluka(Taufkirchen, Germany). Leflunomide and A77 1726 were kindlyprovided by Aventis Pharma (Frankfurt/Main, Germany).

�� 2 Samples were prepared by adding 0.5 mL0.1 mol/L sodium acetate buffer (pH 5) and 0.1 mL internalstandard stock solution (25 �g warfarin/mL distilled water, storedat 4°C) to 0.25 mL plasma followed by the addition of 10 mL ethylacetate. The tubes were capped, shaken for 15 min and centrifugedat 4000 rpm for 5 min. The organic layer was transferred into aglass tube and evaporated to dryness under a gentle nitrogen streamat room temperature. Prior to analysis, the residue was dissolved in200 �L of a solution consisting of acetonitrile, distilled water andformic acid (40/59.8/0.2, v/v). The reconstituted specimens werevortexed for 60 s and the resulting extract was used for HPLCanalysis. The injection volume of the samples was 50 �L. Plasmacalibration standards (0.05, 0.1, 0.2, 0.5, 1, 5, 10, 20, 40, 60, 80,100 �g/mL) were prepared in the same manner. To achievedifferent final concentrations, plasma was spiked with increasingamounts of leflunomide and A77 1726, respectively. For prepara-tion of the standard solution, leflunomide was dissolved in ethanoland 0.9% (m/v) saline (95/5, v/v). The stock standard of A77 1726was prepared in distilled water containing 0.005% (v/v) saturatedsodium hydroxide. Both stock solutions were stored at �80°C forup to one month. The 0.1 mol/L acetate buffer (pH 5) was preparedby dissolving 4.1 g of sodium acetate trihydrate in 500 mL distilledwater and adjusted with acetic acid to a final pH of 5.

/� �� 2 For chroma-tography and detection of the analytes a Jasco HPLC apparatus(Jasco, Labor- und Datentechnik GmbH, Groß-Umstadt, Germany)consisting of a gradient pump (model PU-1585), a variable

wavelength detector (model UV-1575) and a three-line-degasser(model DG-980-50) was used. The column was heated using apeltier-based column heater (Chemdata, Labor und Datentechnik,Groß-Zimmern, Germany). Separation of the analytes wasaccomplished using a reversed-phase column (CC125/3 Nucleosil100-5 C18; Machery-Nagel, Duren, Germany) and a C18 precolumninsert. The column temperature was maintained at 21°C. Themobile phase consisted of acetonitrile, distilled water and formicacid (40/59.8/0.2, v/v). Elution of A77 1726, leflunomide andwarfarin was achieved at a flow rate of 0.5 mL/min. UV detectionwas set at 261 nm. Quantification of peaks was achieved by theinternal standard peak-area ratio method and linear regression.

3�� %41/ ��2 The validation of the methodwas performed according to Shah et al. (2000). The specificity ofthe method was examined by analysing six plasma samples fromdifferent probands who did not take any medication before bloodsampling. Linearity of the standard curves was determined over0.05–100 �g/mL (leflunomide) and 0.1–100 �g/mL (A77 1726),respectively. The standard curves were divided into concentrationranges from the lower limit of quantification (LLOQ) up to 5 �g/mL and from 5 �g/mL up to 100 �g/mL. The intra-day reprodu-cibility of the method was tested by multiple analysis of individualhuman plasma samples on the same day. Inter-day reproducibilitywas assessed on three different days. Stability of analytes wasevaluated and monitored following a one-month period byinjecting standards prepared from the same stock solutions storedat �80°C.

5��6�� 2 For quantification of theanalytes a standard curve for leflunomide and A77 1726 over thevalidated range was generated for each analytical run. Further-more, quality-control samples of five concentrations for eachanalyte (0.1, 1, 5, 40 and 100 �g/mL for A77 1726; 0.05, 1, 5, 40and 100 �g/mL for leflunomide) were incorporated in duplicatesinto each run. The results of the quality-control samples providedthe basis of accepting or rejecting the run.

+0�.1*�

3��

Using the described analytical method, an optimalresolution of the analytes was achieved. Fig. 2 showstypical chromatograms for blank human plasma [Fig.2(A)], human plasma spiked with leflunomide and A771726 at 20 �g/mL [Fig. 2(B)] and plasma from a patientorally dosed with 10 mg leflunomide for 278 days [Fig.2(C)]. The resulting retention times were 8.2 (A77 1726),12.2 (internal standard warfarin) and 16.2 min (lefluno-mide), respectively. The respective peaks were resolvedwith no interfering peaks.

Linearity of the calibration curves for the twocompounds was achieved for concentrations between0.05 and 100 �g/mL (leflunomide) and between 0.1 and100 �g/mL (A77 1726), respectively. In both cases thecorrelation coefficients were greater than 0.995. Theresults of the intra- and inter-day validation of the two

Copyright 2003 John Wiley & Sons, Ltd. Biomed. Chromatogr. 17: 276–281 (2003)

HPLC analysis of leflunomide and its metabolite A77 1726 ORIGINAL RESEARCH 277

analytes are shown in Tables 1 and 2, respectively. Allthree control levels (high, medium and low) for eachcalibration curve used for the validation were consideredacceptable if the coefficient of variation (CV) and theerror of accuracy was less than 15%, with an acceptedtolerance of 20% at the LLOQ. The intra- and inter-assayday coefficients of variation as well as the errors ofaccuracy were within these ranges for all analytes at allthree control levels. Limits of quantification were deter-

mined to be 0.05 �g/mL for leflunomide and 0.1 �g/mLfor A77 1726 in plasma.

Average percentage recoveries of leflunomide and A771726 were assessed at three concentrations and are shownin Table 3. The mean recovery, as measured for the high,medium and low controls relative to the unextractedbuffer controls, was 90–96% (leflunomide) and 85–90%(A77 1726), respectively.

Assay checks carried out over one month with three

Figure 2. Chromatograms of (A) blank human plasma, (B) human plasmaspiked with leflunomide and A77 1726 (at 20 �g/mL each), and (C) plasmaof a patient with rheumatoid arthritis orally dosed with 10 mg leflunomideper day for 278 days. Absorbance wavelength was 261 nm.


278 ORIGINAL RESEARCH A. Schmidt et al.

concentrations of each analyte demonstrated that leflu-nomide and A77 1726 were stable in plasma adjusted to afinal pH of 2 with hydrochloric acid. The differencebetween the measured values before and after a one-month freezing was less than 5% in each case (data notshown). When non-acidified plasma was stored at �80°Cover the same time period, a substantial conversion ofleflunomide to A77 1726 up to 40% of the addedconcentration was observed (data not shown).

4��

To apply the developed and validated method, concen-trations of leflunomide and A77 1726 were assessed inplasma of patients with rheumatoid arthritis treated withdaily doses of 10 or 20 mg leflunomide. In all plasmasamples available, leflunomide was not detectable.Plasma levels of A77 1726 showed a high variabilitywith concentrations ranging from a minimum of 5 �g/mLup to a maximum of 92.7 �g/mL (Table 4). Averageplasma concentrations of A77 1726 were 19.8 � 11.4(patients treated with 10 mg leflunomide per day;mean � SD) and 46.2 � 30.6 �g/mL (patients treated

with 20 mg leflunomide per day; mean � SD), respec-tively (Table 4).

-(�/.��(,)

The present paper describes a sensitive analytical methodfor the simultaneous determination of leflunomide and itsmajor metabolite, A77 1726, in human plasma. Amongvarious isocratic mobile phases tested, the acetonitrile–water phase used with 0.2% formic acid was found to bemost effective in achieving separation of the peaks.

Table 1. Intra- and inter-day precision (CV) and accuracy for low, medium and high concentrations of leflunomide [thestandard curve was split in two ranges (0.05–5 �g/mL and 5–100 �g/mL) and was validated for each range separately]

Added concentra-tion (�g/mL)

Leflunomide Intra-day (n = 5 each) Leflunomide Inter-day (three days, n = 5 each)

Measured concentra-tion (mean � SD,

�g/mL) CV (%) Accuracy (%)



0.05 0.053 � 0.004 7.2 �9.3 0.052 � 0.003 6.2 4.91 0.93 � 0.03 3.5 �6.5 0.94 � 0.03 2.9 �5.65 4.5 � 0.2 3.5 5.3 4.6 � 0.2 4.4 �7.85 5.7 � 0.1 1.8 14.4 5.4 � 0.5 9.6 8.3

40 42.0 � 2.6 6.1 5.0 38.8 � 3.0 7.7 �2.9100 104.3 � 12.2 11.7 4.3 105.2 � 8.2 7.8 5.2

Table 2. Intra- and inter-day precision (CV) and accuracy for low, medium and high concentrations of the leflunomidemetabolite A77 1726 in human plasma [the standard curve was split in two ranges (0.1–5 �g/mL and 5–100 �g/mL) and wasvalidated for each range separately]

Added concentra-tion (�g/mL)

A77 1726 Intra-day (n = 5 each) A77 1726 Inter-day (three days, n = 5 each)





0.1 0.113 � 0.001 1.0 13.2 0.097 � 0.012 12.5 �2.61 0.98 � 0.02 2.4 �1.6 1.0 � 0.02 2.3 �0.15 4.9 � 0.1 2.2 �1.1 5.0 � 0.2 4.3 0.015 5.2 � 0.2 3.8 3.1 5.2 � 0.3 5.6 3.3

40 41.2 � 1.4 3.3 3.0 40.6 � 1.6 3.9 1.6100 94.5 � 3.2 3.4 �5.5 96.5 � 5.9 6.1 �3.5

Table 3. Percentage recovery of leflunomide and A77 1726from human plasma

Added concentration (�g/mL)

Recovery (%)

Leflunomide A77 1726

1 90.5 � 1.4 85.3 � 2.110 91.5 � 1.9 90.2 � 4.5

100 96.2 � 0.6 89.8 � 3.4

Data are means � SD (n = 5 for each group). The percentage recoverywas calculated as (area counts from extracted samples/area countsfrom unextracted samples) � 100. The extraction volume was0.25 mL.



Specificity of the analytical method was indicated by theabsence of interfering peaks as evaluated by chromato-grams of blank human plasma and plasma spiked withleflunomide, A77 1726 and warfarin (internal standard).Linearity was obtained over concentration ranges 0.1–100 �g/mL (A77 1726) and 0.05–100 �g/mL (lefluno-mide). An acceptable and reliable precision and accuracyof the method was demonstrated in intra- and inter-dayvalidation assays. For both analytes, precision andaccuracy fulfilled the validation criteria published byShah et al. (2000) with LLOQ of 0.05 �g/mL (lefluno-mide) and 0.1 �g/mL (A77 1726), respectively. More-over, determination of the stability of the analytesfollowing a one-month-freezing period showed that thesubstances were fairly stable, and did not undergosignificant degradation. To prevent hydrolysis of theisoxazol ring of leflunomide, plasma samples wereadjusted with hydrochloric acid to a final pH of 2 justbefore freezing.

In the past several HPLC methods have been publishedfor the single determination of A77 1726 (Dias et al.,1995; Lucien et al., 1995; Beaman et al., 2002). How-ever, in the present HPLC assay the parent compoundleflunomide and its active metabolite, A77 1726, weredetermined simultaneously. This is a clear advantagecompared to previously published methods. Notably, inone study authors tried to analyze both A77 1726 andleflunomide, but did not consider the hydrolysis ofleflunomide in extraction solvent containing potassiumcarbonate (Dias et al., 1995). In the present study,however, plasma samples were extracted under weakacidic conditions using a solvent comprising of ethylacetate and sodium acetate buffer (pH 5) that preventsdegradation of leflunomide during extraction. Using thisprocedure, acceptable recoveries for both analytes wereobtained.

A77 1726 displays a long elimination half life of

approximately 15 days ranging from 5 to 40 days(Bartlett et al., 1996). To facilitate the rapid attainmentof steady state levels of A77 1726, a loading dose of300 mg leflunomide (given as single 100 mg doses onthree consecutive days) may be administered at thebeginning of the therapy, followed by a maintenance doseof 10 or 20 mg per day. During long-term treatmentwithout a loading dose, steady-state concentrations ofA77 1726 are usually reached within nearly 2 months(Fachinformation Arava�-Filmtabletten, Aventis Phar-ma, Frankfurt/Main, Germany, 2002). In the presentstudy, evidence for method applicability was obtained byassessing the plasma levels of leflunomide and A77 1726at steady state in 12 patients with rheumatoid arthritis.Steady-state concentrations of A77 1726 measured withthe present method (19.8 � 11.4 and 46.2 � 30.6 �g/mLat 10 and 20 mg leflunomide/day, respectively; means �SD) were in good line with steady-state concentrationsfor patients with rheumatoid arthritis reported by themanufacturer. In the above study plasma concentrationsof A77 1726 following treatment with a maintenancedose over 24 weeks were as follows: 18 � 9.6 (n = 18)and 63 � 36 �g/mL (n = 18) at 10 and 25 mg lefluno-mide/day, respectively (means � SD) (FDA, 1998).Plasma concentrations of A77 1726 measured in ourstudy were within the range of the assay limits, suggest-ing that the developed method is sufficient for pharma-cological investigations.

The specific route and place of bioactivation to A771726 is presently unknown. However, a substantialconversion of leflunomide to A77 1726 is supposed totake place already in the gut wall followed by a furtherhydrolysis in the liver (Fachinformation Arava�-Filmta-bletten, Aventis Pharma, Frankfurt/Main, Germany,2002). This assumption was corroborated by the presentstudy showing that leflunomide was not detectable inplasma of patients after administration of a daily 10 or

Table 4. Steady-state plasma concentrations of A77 1726 in patients with rheumatoid arthritis treated with leflunomide at10 mg or 20 mg/day

PatientLeflunomide maintenance

dose (mg)Treatment (days)

before blood samplingPlasma concentration of

A77 1626 (�g/mL)Plasma concentration of A771726 (�g/mL, mean � SD)

1 10 74 34.2 19.8 � 11.42 10 127 17.93 10 256 5.04 10 351 32.35 10 69 17.26 10 278 12.07 20 90 15.2 46.2 � 30.68 20 113 59.19 20 61 18.3

10 20 738 28.311 20 427 63.912 20 817 92.7

Patients received a maintenance dose of 10 or 20 mg leflunomide per day. Blood samples for determination of A77 1726 plasma levels were obtainedfrom several rheumatologist’s practices.


280 ORIGINAL RESEARCH A. Schmidt et al.

20 mg dose of leflunomide. Moreover, these datastrengthens the notion that the pharmacological activityof leflunomide rests in its major metabolite, A77 1726.However, owing to its ability to analyze both A77 1726and leflunomide, our method may offer new opportunitiesfor pharmacokinetic and metabolism studies to assess thespecific site of bioactivation of leflunomide in differenttissues and fluids. Apart from pharmacokinetic studies, asimultaneous determination of A77 1726 and leflunomidemay be used for stability checks of leflunomide-contain-ing pharmaceuticals. Moreover, our method could beapplied routinely for monitoring A77 1726 in humanplasma of patients under leflunomide therapy. Forinstance, a determination of plasma concentrations ofA77 1726 could be performed when leflunomide-treatedpatients develop severe adverse drug reactions (egelevation of hepatic transaminases).

In summary, a reliable and relatively simple analyticalmethod for the simultaneous determination of lefluno-mide and its main metabolite, A77 1726, has beendeveloped and validated. Moreover, the assay presentedhas been found to be well suited for the determination ofleflunomide and A77 1726 plasma concentrations inpatients orally dosed with leflunomide at does currentlyused in clinical practise. The main advantage of themethod is the ability to analyze both compoundssimultaneously under similar conditions, thereby savingtime and expenses for sample preparation.

+070+0)/0�

Bartlett RR, Dimitrijevic M, Mattar T, Zielinski T, Germann T, RudeE, Thoenes GH, Kuchle CC, Schorlemmer HU, Bremer E, FinneganA and Schleyerbach R. Leflunomide (HWA 486), a novelimmunomodulating compound for treatment of autoimmunedisorders and reactions leading to transplantation rejection. Agentsand Actions 1991; 32: 10–21.

Bartlett RR, Campion P, Musikic P, Schleyerbach R, Zielinski T andSchorlemmer HU. Nonsteroidal Anti-inflammatory Drugs: Mecha-nisms and Clinical Uses, Lewis AJ, Furst DE (eds). Marcel Dekker:New York, 1994; 349–366.

Bartlett RR, Brendel S, Zielinski T and Schorlemmer HU. Lefuno-mide, an immunorestoring drug for the therapy of autoimmunedisorders, especially rheumatoid arthritis. Transplantation Proceed-ings 1996; 28: 3074–3078.

Beaman JM, Hackett LP, Luxton G and Illett KF. Effect ofhemodialysis on leflunomide plasma concentrations. Annals ofPharmacotherapy 2002; 36: 75–77.

Breedveld FC and Dayer JM. Leflunomide: mode of action in thetreatment of rheumatoid arthritis. Annals of Rheumatic Disease2000; 59: 841–849.

Cherwinski HM, Cohn RG, Cheung P, Webster DJ, Xu YZ, CaulfieldJP, Young JM, Nakano G and Ransom JT. The immunosuppressantleflunomide inhibits lymphocyte proliferation by inhibiting pyrimi-dine biosynthesis. Journal of Pharmacology and ExperimentalTherapeutics 1995; 275: 1043–1049.

Chong AS, Finnegan A, Jiang X, Gebel H, Sankary HN, Foster P andWilliams JW. Leflunomide, a novel immunosuppressive agent. Themechanism of inhibition of T cell proliferation. Transplantation1993; 55: 1361–1366.

Davis JP, Cain GA, Pitts WJ, Magolda RL and Copeland RA. Theimmunosuppressive metabolite of leflunomide is a potent inhibitorof human dihydroorotate dehydrogenase. Biochemistry 1996; 35:1270–1273.

Dias VC, Lucien J, LeGatt DF and Yatscoff RW. Measurement of theactive leflunomide metabolite (A77 1726) by reverse-phase high-performance liquid chromatography. Therapentic Drug Monitoring1995; 17: 84–88.

FDA. Study YU#204. Clinical pharmacology/biopharmaceutics re-view. Arava� Leflunomide tablets; summary basis of approvalequivalent. FDA Summary Basis of Approval. NDA: 20-905, 1998.

Green S, Watanabe K, Braatz-Trulson J and Lou L. Inhibition ofdihydroorotate dehydrogenase by the immunosuppressive agentleflunomide. Biochemical Pharmacology 1995; 50: 861–867.

Lucien J, Dias VC, LeGatt DF and Yatscoff RW. Blood distributionand single-dose pharmacokinetics of leflunomide. Therapeutic DrugMonitoring 1995; 17: 454–459.

Mattar T, Kochhar K, Bartlett RR, Bremer EG and Finnegan A.Inhibition of the epidermal growth factor receptor tyrosine kinaseactivity by leflunomide. FEBS Letters 1993; 334: 161–164.

Shah VP, Midha KK, Findlay JW, Hill HM, Hulse JD, McGilveray IJ,McKay G, Miller KJ, Patnaik RN, Powell ML, Tonelli A,Viswanathan CT and Yacobi A. Bioanalytical method valida-tion—a revisit with a decade of progress. Pharmaceutical Research2000; 17: 1551–1557.

Silva HT Jr, Cao W, Shorthouse RA, Loffler M and Morris RE. In vitroand in vivo effects of leflunomide, brequinar and cyclosporine onpyrimidine biosynthesis. Transplantation Proceedings 1997; 29:1292–1293.

Xu X, Blinder L, Shen J, Gong H, Finnegan A, Williams JW andChong AS. In vivo mechanism by which leflunomide controlslymphoproliferative and autoimmune disease in Mrl/MpJ-lpr/lprmice. Journal of Immunology 1997; 159: 167–174.



simultaneous determination of leflunomide and its active metabolite, a77 1726, in human plasma by...

Documents