Separation of etodolac enantiomers by capillary electrophoresis. Validation and application of the chiral method to the analysis of commercial formulations

Raquel Rod íguez de Pablos Ca men Ga cía-Ruiz An onio L. C ego M. Luisa Ma ina Depa amen o de Química Analí ica, Facul ad de Química, Uni e sidad de Alcalá, Alcalá de Hena es (Mad id), Spain Sepa a ion o e odolac enan iome s by capilla y elec opho esis. Valida ion and applica ion o he chi al me hod o he analysis o comme cial o mula ions Sepa a ion o e odolac enan iome s, which exhibi di e en biological ac i i y and pha macokine ic p o iles, has been achie ed using he andomly subs i u ed (2-hyd oxy)p opyl-b-cyclodex in (HP-b-CD) as chi al selec o in capilla y elec opho- esis. The selec ion o his CD was made a e sc eening o di e en CD de i a i es o neu al and anionic na u e. The e ec on he enan io esolu ion o he bu e con- cen a ion and o he deg ee o subs i u ion (DS) and concen a ion o he CD as well as o ins umen al pa ame e s, such as he capilla y empe a u e and he sepa a ion ol age, we e s udied. The highes esolu ion o e odolac enan iome s was a ound 2.5 using 100 mMphospha e bu e (pH 7.0) wi h 20 mMHP-b-CD (DS ,4.2) and UV de ec ion a 225 (10) nm wi h a e e ence wa eleng h a 360 (50) nm. Valida ion o he chi al me hod in e ms o selec i i y, linea i y, p ecision (ins umen al epea abili y, me hod epea abili y, in e media e p ecision), and he limi s o de ec ion and quan i a- ion allowed o e alua e i s quali y o he analysis o e odolac enan iome s in di e en pha maceu ical p epa a ions con aining acemic e odolac. Keywo ds: Capilla y elec opho esis / Chi al sepa a ion / Cyclodex ins / Enan iome s / E odolac / (2-Hyd oxy)p opyl-b-cyclodex in DOI 10.1002/elps.200410097 1 In oduc ion E odolac (Fig. 1) is a nons e oidal an i-in lamma o y d ug (NSAID) ha exhibi s an i-in lamma o y, analgesic, and an ipy e ic ac i i ies in animal models. This d ug is ma ke ed as a acemic mix u e o e odolac in which he S- o m is biologically ac i e [1]. Howe e , i has been epo ed e y ecen ly ha he R-enan iome exhibi s pha macological ac i i y in pa ien s su e ing om leuke- mia (B-cell ch onic lymphocy ic leukemia) [2]. Addi ional- ly, he mo e ex ensi e binding o S-e odolac compa ed o he R- o m o human se um albumin unde physiological condi ions has also been epo ed [3, 4]. On he o he hand, enan ioselec i i y was e lec ed in he pha macoki- Figu e 1. Basic s uc u e o e o- dolac ((6)-1,8-die hyl-1,3,4,9- e ahyd opy ano-[3,4-b]indole- 1-ace ic acid); (*) indica es he chi al cen e . ne ics o e odolac enan iome s since he plasma con- cen a ions o he R-enan iome g ea ly exceeded hose o he S- o m [5, 6]. Se e al me hods o he sepa a ion o e odolac enan io- me s by high-pe o mance liquid ch oma og aphy (HPLC) ha e been epo ed showing he in e es o he de elop- men o ch ial me hods o he analysis o e odolac [7–10]. Howe e , he capabili y o capilla y elec opho esis (CE) as sepa a ion echnique o d ug enan iome s is widely ecognized nowadays. Hund eds o scien i ic pape s, a wide numbe o e iews, se e al dedica ed issues o in e na ional jou nals [11–15] and e en a book [16] ha e been published epo ing a la ge numbe o enan iome ic sepa a ions o d ugs by CE. Howe e , un il now, e y ew me hods o he baseline sepa a ion o e odolac enan io- me s ha e been epo ed. Using ancomycin as chi al selec o , Fanali e al. [17] achie ed he sepa a ion o he enan iome s o e odolac in abou 12 min and de ec ed Co espondence: D . M. Luisa Ma ina, Depa amen o de Quí- mica Analí ica, Facul ad de Química, Uni e sidad de Alcalá, C a. Mad id Ba celona Km. 33.600, E-28871 Alcalá de Hena es (Mad id), Spain E-mail: [email p o ec ed] Fax: 134-91-8854971 Abb e ia ions: Ac-â-CD, ace yla ed-b-cyclodex in; ANOVA, analysis o a iance; CE-â-CD, ca bocye hyla ed b-CD; CM-â- CD, ca boxyme hyla ed b-CD; DS, deg ee o subs i u ion; HBen-â-CD, (2-hyd oxy)-bu en-3-yl-b-CD; HP-â-CD, 2-hy- d oxyp opyla ed b-CD; RAMEG, andomly me hyla ed b-CD; Succ-â-CD, succinyla ed b-CD 1106 Elec opho esis 2005, 26, 1106–1113 2005 WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim CE and CEC Elec opho esis 2005, 26, 1106–1113 CE sepa a ion o e odolac enan iome s 1107 hem by mass spec ome y wi h elec osp ay ioniza ion. Vancomycin is a glycopep ide- ype mac ocyclic an ibio ic wi h a limi ed pH wo king ange (4 ,pH ,8), adso bs o he wall o un ea ed used-silica capilla ies and has s ong abso p ion in he UV-Vis egion. On he o he hand, a pe me hyla ed g-CD was used as chi al selec o by Maye and Schu ig [18] in capilla y elec och oma o- g aphy. They showed he use ulness o open- ubula columns o he enan iosepa a ion o NSAIDs including e odolac. Howe e , only wi h he pe me hyla ed g-CD (Chi asil-Dex) i was possible o achie e a esolu ion o 1.6 o e odolac enan iome s in 30 min. Ou goal was o de elop a chi al sepa a ion me hod by CE using cyclodex ins (CDs) as chi al selec o s due o hei po en ial o chi al sepa a ions, comme cial a ailabili y, s abili y, and easy use in compa ison wi h o he chi al selec o s as, o example, mac ocyclic an ibio ics o p o- eins, which p esen simila p ac ical p oblems. Then, in his wo k, we ha e made a sc eening o di e en CD de i a i es o neu al and anionic na u e o selec he mos app op ia e chi al selec o o he enan io esolu ion o e odolac by CE. Wi h he slec ed chi al selec o , e ec on he enan io esolu ion o he bu e concen a ion, deg ee o subs i u ion (comme cial o igin) o he chi al selec o , and concen a ion o he CD, as well as ins u- men al pa ame e s, such as empe a u e and he sepa a ion ol age, ha e been s udied o achie e he bes esolu ion o e odolac enan iome s. Finally, he alida ion o he chi al me hod and i s use o quan i y he amoun o e odolac in di e en pha maceu ical p epa a ions ha e also been pe o med. 2 Ma e ials and me hods 2.1 Reagen s and samples All eagen s employed o he p epa a ion o he sepa a- ion bu e s we e o analy ical g ade. Disodium hyd ogen phospha e hyd a e and sodium ace a e we e supplied by Pan eac (Ba celona, Spain). Dime hyl sul oxide (DMSO) was om Scha lab (Ba celona, Spain). Sodium hyd oxide, sodium dihyd ogen phospha e dihyd a e, and o ophospho ic acid we e om Me ck (Da ms ad , Ge - many). b-Cyclodex in (b-CD), andomly me hyla ed b-cyclodex in (RAMEB, deg ee o subs i u ion (DS), a e - age numbe o subs i uen s on one cyclodex in ing ,12–14), 2-hyd oxyp opyla ed-b-cyclodex in (HP-b- CD, DS ,4.2), ca boxyme hyla ed b-cyclodex in (CM-b- CD, DS ,3), and succinyla ed b-cyclodex in (Succ-b- CD, DS ,3) we e pu chased om Fluka (Buchs, Swi ze - land). (2-Hyd oxy)-bu en-3-yl-b-cyclodex in (HBen-b- CD,DS,5), ca boxye hyla ed-b-cyclodex in (CE-b- CD, DS ,3), ca boxye hyla ed g-cyclodex in (CE-g-CD, DS ,4), ace yla ed b-cyclodex in (Ac-b-CD, DS ,7), ace yla ed g-cyclodex in (Ac-g-CD, DS ,7), ca box- yme hyla ed g-cyclodex in (CM-g-CD, DS ,3), g-cyclo- dex in (g-CD), and HP-b-CD (DS ,3) we e om Cyclolab (Budapes , Hunga y). In addi ion, HP-b-CD (DS ,4.7) and RAMEB (DS ,10.5–14.7) we e also ob ained om Sigma (Mad id, Spain). Wa e used o p epa e solu ions was pu- i ied wi h a Milli-Q sys em om Millipo e (Bed o d, MA, USA). All solu ions we e il e ed p io o use h ough 0.45 mm po e size disposable nylon il e s om Sugelabo (Mad id, Spain). E odolac was om Sigma. The pha ma- ceu ical able s analyzed we e ob ained in a pha macy o Ismi a (Tu key). Table 1 shows he composi ion o hese o mula ions. Table 1. Composi ion o he h ee di e en comme cial o mula ions analyzed Comme cial o mula ion Composi ion Table A E odolac (200 mg pe able ), Fe2O3and TiO2 Table B E odolac (300 mg pe able ), Fe2O3, FeO, and TiO2 Table C E odolac (300 mg pe able ), Fe2O3and TiO2 2.2 Appa a us All expe imen s we e pe o med on a HP3D CE sys em (Hewle -Packa d, Waldb onn, Ge many) equipped wi h an on-column diode a ay de ec o (DAD). Ins umen con ol and da a acquisi ion was pe o med wi h he HP 3D-CE ChemS a ion so wa e. Sepa a ions we e pe - o med on uncoa ed used-silica capilla ies o 50 mmID and 375 mm OD, pu chased om Composi e Me al Se - ices (Wo ces e , England). Capilla ies had a o al leng h o 58.5 cm and 50 cm o he de ec o . The capilla y em- pe a u e was 157C and UV de ec ion was pe o med a 225 nm wi h a bandwid h o 10 nm using a e e ence wa eleng h o 360 nm wi h a bandwid h o 50 nm and a esponse ime o 0.3 s. To adjus he pH o he sepa a ion bu e s, a 744 pH-me e om Me ohm (He isau, Swi ze - land) was used. 2.3 P ocedu es Be o e i s i s use, a new capilla y was lushed (1 ba ) wi h 1MNaOH o 30 min, ollowed by 30 min wi h wa e and 60 min wi h bu e . Be ween in oduc ion o samples, he capilla y was condi ioned wi h 0.1 MNaOH o 1 min ol- lowed by he sepa a ion bu e o 2 min. Injec ions we e made by p essu e (50 mba o 3 s o sample ollowed by 50 mba o 3 s o bu e ), and he applied ol age was 2005 WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim 1108 R. R. de Pablos e al. Elec opho esis 2005, 26, 1106–1113 20 kV. Bu e sepa a ion solu ions we e p epa ed by dis- sol ing he same amoun o he acid and basic compo- nen s in o de o ge he desi ed pH, he pKao he selec ed acid. The app op ia e amoun o each CD was weighed and dissol ed in o hese bu e solu ions a a inal concen a ion o 20 mM. S ock solu ions we e p e- pa ed by dissol ing he app op ia e amoun o e odolac in DMSO o achie e a inal concen a ion o 2000 mg/L. These solu ions we e dilu ed in 1:1 / wa e /DMSO o ge dilu ed solu ions wi h concen a ions o e odolac be- ween 10 and 200 mg/L (5 and 100 mg/L o each enan- iome ). Sample solu ions o he h ee comme cial o - mula ions analyzed we e p epa ed by c ushing and mix- ing homogeneously i e able s o each comme cial o mula ion and weighing abou 0.1 g (0.11–0.13 g) o he esul an powde which was dissol ed in DMSO a a inal concen a ion o 5000 mg/L. These sample solu ions we e dilu ed 1:1 / in wa e /DMSO o ge solu ion con- cen a ions o 100 mg/L. All solu ions (bu e s, s anda ds, and samples) we e il e ed h ough 0.45 mm po e size nylon il e memb anes and sonica ed be o e use. 2.4 Da a ea men Resolu ion was ob ained om he ChemS a ion so wa e using he equa ion Rs = 1.18 22 1/w1/2, 11w1/2, 2 whe e 1 and 2a e he mig a ion imes o e odolac enan iome s and w1/2,1 ,w1/2,2 a e hei peak wid hs a hal heigh . Lim- i s o de ec ion (LODs; 3sa/b) and quan i a ion (LOQs, 10sa/b) we e de e mined om he s anda d e o o he in e cep (sa) and he slope (b) o he calib a ion cu e ob ained by analysis o a iance (ANOVA) [19]. Con ol g aphs elabo a ed o s abili y s udies we e made plo ing co ec ed peak a eas as a unc ion o ime. The cen al alues, X, we e calcula ed as he a e aged alue o six epea ed uns, and he con ol limi s (which enable o isualize he alues in and ou o con ol) we e calcula ed as X63swhe e sis he s anda d de ia ion o he six eplica e alues [20]. Expe imen al da a analysis and pa- ame e s we e calcula ed using Excel Mic oso XPand S a g aphics PlusVe sion 5.0 so wa es. 3 Resul s and discussion 3.1 Me hod op imiza ion The c ucial s ep o achie e an enan iome ic sepa a ion is he choice o he chi al selec o . CDs a e he a ou i e chi al selec o s in CE and hey ha e shown o sepa a e he enan iome s o a huge amoun o chi al d ugs. Due o hei a ailabili y, s abili y, and po en ial, we selec ed di - e en CD de i a i es o neu al and anionic na u e (see Table 2) o sc eening hei capabili y as chi al selec o s o e odolac enan iome s. All hese CDs we e used in 50 mM phospha e bu e a pH 7.0. This bu e was used because i enabled he la ge di e ence be ween he EOF and e o- dolac signals ( he pKao e odolac is 4.65). In addi ion, he bu e pH is simila o physiological condi ions which is desi able o pha macokine ic s udies whe e biological luids a his pH (e.g., blood) may be analyzed. As shown in Table 2, he anionic cyclodex ins CM-g-CD, CM-b-CD, CE-g-CD, CE-b-CD, and Succ-b-CD and he neu al cyclodex ins b-CD, RAMEB om Sigma, Ac-b- CD, and Ac-g-CD (all o hem a a concen a ion o 20 mM) did no show enan io esolu ion powe o e odolac. Wi h 20 mMRAMEB om Fluka o HBen-b-CDpoo enan io- esolu ions we e obse ed (Rs ,0.7). The highes enan- io esolu ions we e ob ained o he neu al HP-b-CD and g-CD a a concen a ion o 20 mMin phospha e bu e a pH 7.0 (see Table 2). As a consequence, we selec ed HP- b-CD as chi al selec o o he sepa a ion o e odolac enan iome s because i is less expensi e han g-CD. Table 2. Enan io esolu ions (Rs) o e odolac enan iome s ob ained wi h he di e en CD de i a i es s ud- ied a a 20 mMconcen a ion CD (comme cial supplie ) Enan io esolu ion Neu al CDs b-CD (Fluka) 2 RAMEB (Sigma) 2 RAMEB (Fluka) 1 HP-b-CD (Cyclolab) 11 HBen-b-CD(Cyclolab) 1 Ac-b-CD (Cyclolab) 2 g-CD (Cyclolab) 11 Ac-g-CD (Cyclolab) 2 Anionic CDs Succ-b-CD (Fluka) 2 CM-b-CD (Fluka) 2 CE-b-CD (Cyclolab) 2 CM-g-CD (Cyclolab) 2 CE-g-CD (Cyclolab) 2 2,Rs,0; 1,Rs,0.7; 11, 0.7 Rs1 Wi h he selec ed CD, he in luence on enan iome ic es- olu ion o he empe a u e and sepa a ion ol age was s udied o choose he mos app op ia e alues o hese pa ame e s.The a ia ion o he esolu ion o e odolac enan iome s as a unc ion o he capilla y empe a u e in he ange o 15–457C. showed ha he esolu ion g adu- ally dec eases as he empe a u e inc eases. This e ec was obse ed p e iously wi h he same CD de i a i e by ou esea ch g oup [21] and could be a ibu ed o he inc ease in he s abili y o selec o -selec and complexes 2005 WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim Elec opho esis 2005, 26, 1106–1113 CE sepa a ion o e odolac enan iome s 1109 ha akes place when he empe a u e is dec eased [16, 21, 22]. The e o e, all subsequen sepa a ions we e pe - o med a 157C. Howe e , a low in luence o he sepa a- ion ol age on he enan io esolu ion o d ugs is gene ally ob ained. Fo e odolac enan iome s a simila enan io- esolu ion was ob ained o 15 and 20 kV whe eas he esolu ion dec eased o 30 kV. Since he mig a ion imes dec eased when inc easing he applied ol age, 20 kV o sepa a ion ol age was used in ou expe imen s. The e ec o he bu e concen a ion on he sepa a ion o e odolac enan iome s was in es iga ed by using di e en concen a ions (40, 50, 80, and 100 mM) phospha e bu e con aining 20 mMHP-b-CD. As i can be seen in Fig. 2, he e is a linea inc emen o he enan io esolu ion wi h he phospha e bu e concen a ion om 40 o 100 mM. This can be explained because when he bu e con- cen a ion inc eases he solu e-wall in e ac ions and he undesi able e ec s o elec odispe sion a e minimized [16]. The e o e, 100 mMphospha e bu e a pH 7 (,140 mA) was used o u he expe imen s. The in luence o he concen a ion and he deg ee o subs i u ion o HP-b-CD (comme cial o igin) in he selec ed bu e was also s udied. The comme cial o igin o he chi al selec o in he sepa a ion bu e was ound o be a e y decisi e ac o in luencing he enan iome ic esolu ion o he e odolac enan iome s. This is because Figu e 2. E ec o he bu e concen a ion on he enan- io esolu ion o e odolac o a phospha e bu e (pH 7) wi h 20 mMHP-b-CD (Cyclolab) as chi al selec o . CE condi ions: empe a u e, 157C; un ea ed used-silica capilla y, 58.5 cm (50 cm o he de ec o window)650 mm ID; applied ol age, 20 kV; injec ion, 50 mba 63s.UV de ec ion a 230 nm. HP-b-CD is a andomly subs i u ed CD manu ac u ed in a di e en way by he h ee di e en supplie s which p o ide di e en DSs (see Sec ion 2). Thus, depending on he DS, he subs i u ion pa e ns can be di e en and e en side chain subs i u ion (e.g., 2-(2’-hyd oxy)p opoxy) p opyl subs i uen s) is possible. In addi ion, he DS depends on he chemical eac ion (used base, low/high empe a u e) and la ge di e ences be ween ba ches and p oduc ion si es may be ob ained unde some condi ions. Due o hese p oblems, a end in chi al analysis by CE is he use o single-isome CDs, which a e able o achie e highe ep oducibili y and selec i i y han andomly subs i u ed CDs [23]. Howe e , i is impo an o no e ha HP-b-CD, which has been shown o be he bes chi al selec o o e odolac, is no a single isome (Cyclolab p oduces he single isome monoHp-b-CD bu i s aque- ous solu ili y is e y poo , ,0.05 g/100 cm3, being no use ul in aqueous CE). Fo his eason, we s udied he e ec o he concen a ion o his CD om h ee di e en comme cial supplie s on he esolu ion o e odolac enan- iome s. This s udy was pe o med because he con- cen a ion o he chi al selec o a ec s di ec ly he chi al p ope ies [21, 24]. As can be seen in Fig. 3, when a ying he concen a ion o HP-b-CD o he h ee di e en com- me cial o igins om 10 o 60 mM, he maximum enan io- esolu ion o e odolac was obse ed a a 20 mMCD con- cen a ion. In addi ion, HP-b-CD om Fluka (DS ,4.2) ge s always he bes enan io esolu ion alues o all he concen a ions s udied. As a consequence, his con- cen a ion o he HP-b-CD (DS ,4.2) was chosen o he enan iome ic sepa a ion o e odolac. Figu e 3. In luence o he concen a ion and he DS (comme cial o igin) e ec o he HP-b-CD in 100 mM phospha e bu e (pH 7) on he enan io esolu ion o e o- dolac. CE condi ions as in Fig. 2. 2005 WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim 1110 R. R. de Pablos e al. Elec opho esis 2005, 26, 1106–1113 The de ec ion condi ions o e odolac enan iome s we e also s udied in his wo k. As i is known, he esponse ime o he de ec o a ec s he baseline noise and he enan- io esolu ion. In his case, an inc ease in he esponse ime om 0.1 o 0.3 s leads o a dec ease in he noise ( om 0.7961024 o 0.4361024A.U.) whe eas enan io- esolu ion was main ained (Rs 2.1). Howe e , o a e- sponse ime o 1.0 s al hough he noise dec eased, he da a acqui ed was no enough o show any enan io- esolu ion. As a esul , a esponse ime o 0.3 s was used o a o d he lowes noise (highes signal- o-noise a io, S/N) wi h an adequa e esolu ion. In addi ion, he op imi- za ion o he analy ical wa eleng h alue, he e e ence wa eleng h, and he bandwid h (i will be indica ed wi hin pa en heses) o hose wa eleng hs was pe o med. P e- limina y s udies we e ca ied ou a an analy ical wa e- leng h o 230 nm (4) [18, 25] wi h he lowes possible bandwid h and wi hou e e ence wa eleng h. F om he abso p ion spec um o e odolac i was obse ed ha he abso p ion band o maximum abso p ion is a ound 230 nm. Then 225, 230, and 235 nm we e used as ana- ly ical wa eleng hs ob aining he highes S/N a 225 nm. A his wa eleng h, he bandwid h and he e e ence wa eleng h we e s udied. The lowes noise and highe S/N we e ob ained using 225 nm (10) and a e e ence wa eleng h o 360 nm (50). The elec ophe og am shown in Fig. 4 o he sepa a ion o e odolac enan iome s was ob ained unde hese de ec ion condi ions. A esolu ion .2 o e odolac enan iome s can be obse ed when using 100 mMphospha e bu e (pH 7.0) wi h 20 mMHP-b- CD (DS ,4.2). Figu e 4. Elec ophe og am co esponding o he sepa a ion o e odolac enan iome s using 20 mMHP-b- CD (Fluka) in 100 mMphospha e bu e (pH 7). CE condi- ions: UV de ec ion a 225 (10) nm and e e ence a 360 (50) nm; ime esponse o de ec o , 0.3 s. O he condi ions as in Fig. 2. 3.2 Me hod alida ion P io o he alida ion o he chi al me hod de eloped o e odolac he p econdi ioning o he capilla y and he p o- cedu e o sample injec ion (wi h o wi hou plug) we e s udied. A p econdi ioning wi h he sepa a ion bu e o wi h 0.1 MNaOH ollowed by sepa a ion bu e by p es- su e (1 ba o 2 min) o ol age (2 kV) showed ha a ol - age s age did no imp o e he p ecision (measu ed as RSD o mig a ion imes o six eplica es o a s anda d solu ion). Howe e , he use o NaOH ollowed o back- g ound elec oly e (BGE) ga e he bes p ecision wi hou any loss o enan io esolu ion. A e injec ion o he sample (50 mba o 3 s), a plug (50 mba o 3 s) o DMSO o sepa a ion bu e was injec ed. An imp o emen o he p ecision wi hou loss in enan io esolu ion was obse ed when a plug o BGE a e sample injec ion was used (wi h DMSO a dec ease in he enan io esolu ion and p ecision was ob ained). The p oposed analy ical me hod o he analysis o e o- dolac enan iome s was alida ed by e alua ing he qual- i y, eliabili y, and consis ency o i s esul s. Ne e heless, p io o he alida ion, a s udy on sample s abili y was ca ied ou . E odolac concen a ions o 10 and 100 mg/L (5 and 50 mg/L o each enan iome ) we e p epa ed and hen injec ed e e y day o a o al o i e days. As i can be seen in Fig. 5, he co ec ed peak a eas ob ained o he wo independen samples analyzed (samples A and B) a e included in o he con ol limi s. The alida ion o he chi al CE me hod was made by s udying he p ecision, linea i y, LODs and LOQs, and selec i i y o di e en comme cial pha maceu ical able s. P ecision was e alua ed conside ing he ins u- men al and me hod epea abili y as well as he in e - media e p ecision. Ins umen al epea abili y was de e - mined om six epea ed injec ions o wo s anda d solu- ions a wo di e en concen a ion le els (10 and 100 mg/L e odolac). In bo h cases, RSD alues o co ec ed peak a eas below 5% we e ob ained (Table 3). The me hod epea abili y was assessed wi h six eplica e s anda d solu ions o 10 and 100 mg/L e odolac injec ed by ipli- ca e on he same day. RSD alues o bo h s anda d solu ions we e lowe han 6.7% o co ec ed peak a eas (Table 3). The in e media e p ecision o he me hod was calcula ed o h ee eshly p epa ed s anda d solu ions o e odolac con aining 10, 50, and 100 mg/L and injec ed by iplica e o i e consecu i e days. Table 3 shows ha in his case he RSD alue o co ec ed peak a eas was lowe han 8.2%. The linea i y was s udied plo ing co ec ed peaks a eas (Ac) as a unc ion o analy e concen a ions (c) in mg/L. Fi e s anda d solu ions con aining acemic e odolac om 2005 WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim Elec opho esis 2005, 26, 1106–1113 CE sepa a ion o e odolac enan iome s 1111 Figu e 5. S udy o he sample s abili y wi h 5 and 50 mg/L solu ions o e odolac enan iome s using 20 mMHP-b-CD (DS , 4.2) in 100 mMphospha e bu e (pH 7). Injec ion, sample 50 mba 63 s wi h a plug o BGE a 50 mba 63 s. O he condi- ions as in Fig. 4. Table 3. P ecision o he chi al me hod de eloped o e odolaca) E odolac concen a ion (mg/L) RSD (%) peak a ea Fi s enan iome Second enan iome Ins umen al epea abili y ( n =6) 10 4.1 4.3 100 4.9 4.7 Me hod epea abili y ( n =6) 10 5.8 5.3 100 6.7 5.4 In e media e p ecision ( n = 15) 10 7.2 8.2 50 7.3 7.6 100 8.1 8.2 a) Expe imen al condi ions as in Fig. 4 10 o 200 mg/L (each one injec ed in iplica e) we e con- side ed. The chi al me hod enabled o s udy sepa a ely each enan iome , which concen a ions anged om 5 o 100 mg/L o each enan iome . By ANOVA we con i med ha expe imen al da a i p ope ly o linea models o each one o he enan iome s ( he lack o i was always s a is ically smalle han he pu e e o , p- alue 0.51 and 0.45 o enan iome s). The linea equa ions ob ained we e Ac = 0.0668 c– 0.07 (sa= 0.13; sb= 0.0021; = 0.994) o he i s -mig a ing enan iome and Ac = 0.0648 c– 0.07 (sa= 0.13; sb= 0.0021, = 0.994) o he second-mig a ing enan iome , whe e sais he s anda d e o o he in e cep , sbis he s anda d e o o he slope, and is he co ela ion coe icien . F om calib a ion line pa ame e s, LODs and LOQs o e odolac enan iome s we e calcula ed. Fo bo h enan iome s he LOD was 5.8 mg/L and he LOQ was 19.4 mg/L. An adequa e selec i i y o he analysis o h ee comme - cial p epa a ions was ob ained wi h he chi al me hod de eloped o e odolac. Figu e 8 shows he enan iome ic sepa a ion o e odolac o h ee di e en comme cial able s. I can be obse ed ha he enan io esolu ion o e odolac is good and he e a e no e idence o in e - 2005 WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim 1112 R. R. de Pablos e al. Elec opho esis 2005, 26, 1106–1113 Figu e 6. Elec ophe og am co esponding o he sepa a ion o e odolac enan iome s in h ee di e en comme cial o mula ions, using 20 mMHP-b-CD (DS , 4.2) in 100 mMphospha e bu e (pH 7). CE condi ions as in Fig. 4. e ences due o he o he componen s p esen in he comme cial o mula ion (see Table 1). In addi ion, peaks co esponding o e odolac enan iome s in hese samples we e con i med by compa ison o hei UV abso p ion spec a wi h hose o he s anda d. I should be aken in o accoun ha mig a ion imes o able s a e sho e han hose o s anda d solu ions. This can be explained by he p esence o o he componen s in he pha maceu ical o - mula ions ( ew mino oxides as illus a ed in Table 1) which a ec s he EOF. I is al eady known ha o he componen s, such as cellulose de i a i es, equen ly p esen in able s, a ec he EOF, he eby modi ying he mig a ion imes in a signi ican way [26]. Howe e , his p oblem is easily sol ed using co ec ed mig a ion imes and a eas, as i has been done in his wo k. Finally, Table 4 shows he amoun o e odolac calcula ed o he h ee comme cial o mula ions analyzed. The quan i ica ion o e odolac enan iome s con i ms ha he d ug is p esen in able s as acema e (1:1 mix u e). In addi ion, esul s ob ained show ha he e odolac amoun p esen in able s is inside he ange o he de e mined amoun . The e o e, a good ag eemen be ween he amoun o e odolac de e mined by he chi al CE me hod de eloped and ha decla ed in he label o he pha ma- ceu ical p epa a ion was obse ed. 4 Concluding ema ks In his wo k, he possibili ies o he andomly subs i u ed HP-b-CD as chi al selec o o e odolac a e p esen ed. Al hough di e en enan io esolu ions we e ob ained o his CD wi h di e en DSs depending on he supplie (Cyclolab, Fluka, o Sigma), he HP-b-CD o Fluka (DS , 4.2) allowed o ob ain a ep oducible esolu ion .2 unde he applied expe imen al condi ions (100 mMphospha e bu e , pH 7.0, wi h 20 mMHP-b-CD o he CE sepa a ion, and UV de ec ion a 225 (10) nm wi h e e ence a 360 (50) nm). The chi al CE me hod using he selec ed condi- ions was alida ed o show he quali y, eliabili y, and consis ency o i s esul s. Good pe o mance wi h ega d o linea i y, ins umen al epea abili y, me hod epea - abili y, in e media e p ecision, and selec i i y was achie ed. LODs o 5.8 mg/L and LOQs o 19.4 mg/L we e ob ained o each e odolac enan iome s. The alida ed me hod was applied o de e mine he amoun o e odolac enan iome s in h ee di e en comme cial able s, being sui able o such pu poses. The au ho s hank he Minis y o Science and Technology (Spain) o he esea ch p ojec BQU2003–03638 and he Uni e si y o Alcalá o he esea ch p ojec UAH2002/ 057. C. Ga cía-Ruiz also hanks o he Minis y o Science and Technology o he con ac om he Ramón y Cajal p og am. Au ho s also hank Cyclolab (Budapes , Hun- ga y) o he kind gi o HBen-â-CD, CE-â-CD, and Ac-â-CD used in his wo k and D . Laszlo Jicsinsky o his commen s du ing he elabo a ion o his a icle. Recei ed July 30, 2004 Table 4. A e aged amoun o e odolac enan iome s and o al e odolac measu ed in he h ee di e en comme cial o mula ions analyzeda) Comme cial o mula ion Weigh (mg) i s enan iome 6 s/n1/2 Weigh (mg) second enan iome 6 s/n1/2 To al weigh (mg) 6 s/n1/2 RSD (%) Co ec ed peak a ea To al labeled weigh (mg) Table A 105.2 65.9 105.3 66.2 210.5 612.1 7.6 200 Table B 157.4 68.2 157.3 68.5 314.7 616.8 7.1 300 Table C 156.2 67.5 156.9 68.8 313.1 616.2 6.9 300 a) sis he s anda d de ia ion, = 2.262 ( o a con idence linea ange o 95%), and n=9. 2005 WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim Elec opho esis 2005, 26, 1106–1113 CE sepa a ion o e odolac enan iome s 1113 5 Re e ences [1] The Comp ehensi e Resou ce o Physicians, D ug and Illness In o ma ion, ,h p://www. xmed.com/b.main/b2. pha maceu ical/b2.1.monog aphs/CPS-%20Monog aphs/ CPS-%20(Gene al%20Monog aphs-%20A)/APO_ETODOL AC.h ml., checked on July 2004. [2] Jenson, M., Enge , A., Goebele , M. E., Weissinge , F., Wil- helm, M., Os e bo g, A., Knau , W., Kimby, E. K., Rosen, P. J., Poyn on, C., Rummel, M. J., Hillmen, P., Oli , I., Kennedy, B., Jullusson, G., La en, L., Sc an on, S. A., Pa adiso, L. J., Blood 2003, 102, Abs ac 2496. [3] Ne e , P., Lapicque, F., Benzoni, D., Bannwa h, B., Leloe , X., Kun z, J. L., Schae e beke, T., Be in, P., Gille , P., Dehais, J., T e es, R., D opsy, R., Clin. Pha macol. The . 1993, 53, 152–152. [4] Migno , I., P esle, N., Lapicque, F., Mono , C., D opsy, R., Ne e , P., Chi ali y 1996, 8, 271–280. [5] B ocks, D. R., Jamali, F., Russell, A. S., Skei h, D. J., J. Clin. Pha macol. 1992, 32, 982–989. [6] B ocks, D. R., Jamali, F., Clin. Pha macokine . 1994, 26, 259–274. [7] Ali, I., Aboul-Enein, H. Y., Biomed. Ch oma og . 2003, 17, 113–117. [8] U ay, G., Maie , N. M., Enan iome 1996, 1, 211–217. [9] Caccamese, S., Chi ali y 1993, 5, 164–167. [10] Becke scha enkamp, U., Blaschke, G., J. Ch oma og . B 1993, 621, 199–207. [11] Fanali, S. (Ed.), Elec opho esis 1999, 20, 2577–2800. [12] Fanali, S., Chank e adze, B. (Eds.), Elec opho esis 2001, 22, 3077–3354. [13] Nishi, H., Te abe, S. (Eds.), J. Ch oma og . A 2000, 875,1– 471. [14] W en, S. A. C. (Ed.), Ch oma og aphia 2001, 54, S5–S93. [15] Chank e adze, B., Fanali, S. (Eds.), Elec opho esis 2003, 24, 2407–2732. [16] Chank e adze, B., Capilla y Elec opho esis in Chi al Analy- sis, John Wiley & Sons, Chiches e , UK 1997. [17] Fanali, S., Deside io, C., Schul e, G., Hei meie , S., S ick- mann, D., Chank e adze, B., Blaschke, G., J. Ch oma og . A 1998, 800, 69–76. [18] Maye , S., Schu ig, V., J. Liq. Ch oma og . 1993, 16, 915– 931. [19] Cu ie, L. L. A., Pu e Appl. Chem., 1995, 67, 1699–1723. [20] Mullins, E., Analys 1999, 124, 433–442. [21] Ma ín-Biosca, Y., Ga cía-Ruiz, C., Ma ina, M. L., Elec o- pho esis 2001, 22, 3216–3225. [22] Ve leysen, K., Sand a, P., Elec opho esis 1998 19, 2798– 2833. [23] Amini, A., Elec opho esis 2001, 22, 3107–3130. [24] Beche , I., Pâques, P., Fille , M., Hube , P., C ommen, J., Elec opho esis 1994, 15, 818–823. [25] Maye , S., Schu ig, V., Elec opho esis 1994, 15, 835–841. [26] Ekie , E., Ga cía-Ruiz, C., Ga cía, M. A., Ma ina, M. L., Elec opho esis 2003, 24, 2680–2686. 2005 WILEY-VCH Ve lag GmbH & Co. KGaA, Weinheim