Transcript
1 1 3 ResearchArticle Drug Testingand Analysis Received: 5 September 2009 Revised: 25 October 2009 Accepted: 27 October 2009 Published online in Wiley Interscience: 26 February 2010 (www.drugtestinganalysis.com) DOI 10.1002/dta.92 Spectrophotometricandspectrodensitometricdeterminationoftriamtereneandxipamideinpureformandinpharmaceuticalformulation NourE.Wagieh, a SamahS.Abbas, b M.Abdelkawy b andMahaM.Abdelrahman a ∗ Sensitive and validated UV-spectrophotometric, chemometric and TLC-densitometric methods were developed for determi-nation of triamterene (TRM) and xipamide (XIP) in their binary mixture, formulated for use as a diuretic, without previousseparation. Method A is the isoabsorptive point spectrophotometry, in which TRM concentration alone can be determined atits λ max while XIP concentration can be determined by measuring total concentration of TRM and XIP at their isoabsorptivepointfollowedbysubtraction.MethodBistheratiosubtractionspectrophotometry,whereXIPcanbedeterminedbydividingthe spectrum of the mixture by the spectrum of TRM (as a divisor) followed by subtracting the constant absorbance value of the plateau region, then finally multiplying the produced spectrum by the spectrum of the divisor, while TRM concentrationcan be determinedat its λ max . Method C is a chemometric-assistedspectrophotometry where classical least squares, principalcomponent regression, and partial least squares were applied. Method D is a TLC-densitometry; this method depends onquantitative densitometric separation of thin layer chromatogram of TRM and XIP using silica gel plates at 254nm. TheproposedmethodsweresuccessfullyappliedfortheanalysisofTRMandXIPintheirpharmaceuticalformulationandtheresultswerestatisticallycomparedwiththeestablishedHPLCmethod. Copyright c 2010JohnWiley&Sons,Ltd.Keywords: triamterene; xipamide; isoabsorptive point; ratio subtraction; chemometric; TLC-densitometry Introduction Triamterene (TRM) is 6-phenyl-2,4,7-triaminopteridine. [1] It is a di-uretic drug belonging to the potassium-sparing diuretic family. [2] It is generally administered together with other, more powerful,diuretics such as derivatives of anthranilic acid and thiazide, withthe purpose of reducing their potassium-wasting effects. [2] TRMhas alimited diuretic efficacy. Itacts on the collectingtubules andcollecting ducts, inhibiting sodium reabsorption and decreasingpotassium excretion. After oral administration, TRM is absorbedand metabolized by hydroxylation and subsequent immediateconjugation rendering its main metabolite hydroxytriamterenesulfate. [3] Xipamide (XIP) is 4-chloro-2 ,6 -dimethyl-5-sulfamoylsalicy-lanilide. [1] ithasamoderatelypowerfuldiureticaction.Itdecreasesactive reabsorption of sodium and accompanying chloride bybinding to the chloride site of the electroneutral Na + /Cl − co-transport system and inhibiting its action. [3] The structuralformula of TRM and XIP are shown in Figure 1. The literature survey reveals several analytical methods forthe determination of TRM including spectrophotometric, [4–6] fluorimetric [7–9] and chromatographic [10–13] methods, while XIPwas determined by differentHPLC methods. [14–17] Few methods for the determination of TRM and XIP in theirbinary mixture have been reported in literature, including HPLCmethods [18–22] and GC method. [23] Theaimofthisworkistodevelopsimple,sensitiveandselectiveanalytical methods for the determination of TRM and XIP intheir binary mixture and in pharmaceutical formulation withoutprevious separation. Experimental Instruments AdoublebeamUV-Visiblespectrophotometer(SHIMADZU,Japan)modelUV-1601PCwithquartzcellof1 cmpathlength,connectedto an IBM compatible computer. The software was UVPC personalspectroscopy software v. 3.7. The spectral bandwidth was 2 nmand wavelength-scanning speed 2800nm/min. All data analysiswasperformedusingPLS-Toolbox2.0runningunderMATLAB ® ,v.6.5. [24] A UV lamp with short wavelength 254 nm UV Lamp (ViberLourmat, MAR´NE LA VALLEE Cedex 1, France). A TLC scanner3 densitometer (Camag, Muttenz, Switzerland). The followingrequirements are taken into consideration: slit dimensions:5 × 0 . 2 mm; scanning speed: 20 mm/s; spraying rate: 10 s µ L − 1 ;and data resolution: 100 µ m/step. TLC plates (20 × 20 cm)coated with silica gel 60F 254 (Fluka, Sigma-Aldrich Chemie GmbH,Germany). A sample applicator for TLC Linomat IV with 100 µ Lsyringe (Camag, Muttenz, Switzerland). ∗ Correspondenceto:MahaM.Abdelrahman,AnalyticalChemistryDepartment,FacultyofPharmacy,Beni-SuefUniversity.E-mail:maha m
[email protected] a Analytical Chemistry Department, Faculty of Pharmacy, Beni-Suef University,HussinEl-ShafeiyStreet,57889, Beni-Suef,Egypt. b Analytical Chemistry Department, Faculty of Pharmacy, Cairo University, Kasr El-AiniSt,11562, Cairo,Egypt DrugTest.Analysis 2010 , 2 , 113–121 Copyright c 2010 John Wiley & Sons, Ltd. 1 1 4 Drug Testingand Analysis N. E. Wagieh etal . NNNNH 2 NNH 2 (a)TriamtereneMol. Formula C 12 H 11 N 7 Mol. Wt. 253.27(b)XipamideMol. Formula C 15 H 15 ClN 2 O 4 SMol. Wt. 354.81NHOClSH 2 NOHOOCH 3 H 3 CNH 2 Figure1. ChemicalStructure of Triamterene (a) and Xipamide (b). Materials Purestandards Triamterene and xipamide were kindly supplied by EgyptianInt. Pharmaceutical Industries Co. E.I.P.I. Co. (10th of RamadanCity, Egypt). Their purity was found to be 100.21% and 100.07%,respectively, according to the reported high performance liquidchromatography (HPLC)method. [20] Pharmaceuticaldosageform Epitens ® tablets (Batches No. 078125, 088004, 076677) labelledto contain 30 mg of triamterene and 10 mg of xipamide,manufacturedbyEgyptianInt.PharmaceuticalIndustriesCo.E.I.P.I.Co. (10th of Ramadan City, Egypt). Chemicalsandreagents All chemicals used throughout this work were of analytical grade,andthesolventswereofspectroscopicgrade.Chloroformand33%ammonia solution (El-Nasr Pharmaceutical Chemicals Co., Abu-Zabaal, Cairo, Egypt); methanol of HPLC grade (CHROMASOLV ® ,Sigma-Aldrich Chemie GmbH, Germany). Standardsolutions Stock standard solutions of TRM and XIP (1 mg mL − 1 ): 0.1 gm of TRM and XIP were accurately weighed into two separate 100-mLvolumetric flasks; 50 mL of methanol was added to each flask,shaken to dissolve and then the volume was completed to themark with methanol. Working standard solutions of TRM andXIP (100 µ g mL − 1 ): 10 mL of each of TRM and XIP stock standardsolutions(1 mg mL − 1 )wastransferredaccuratelyintotwoseparate100-mL volumetric flasks; then the volume was completed to themark with methanol. Laboratorypreparedmixtures Mixtures containing different ratios of TRM and XIP wereprepared using their respective working solutions (100 µ g mL − 1 in methanol). Procedures Isoabsorptivepointspectrophotometricmethod Into two separate sets of 10-mL volumetric flasks, differentaliquots containing 20–120 µ g and 20–100 µ g of TRM andXIP, respectively, were accurately transferred from their workingsolutions;thevolumewasthencompletedwithmethanol.Thezeroorder absorbance of each set was recorded and the absorbenciesat λ 225 . 2 nm (Aiso) for XIP and at λ max 367.8nm for TRM weremeasured. The calibration curves relating the absorbance of eachcurve at the selected wavelength to the corresponding drugconcentrations were constructed and the regression equationcorresponding to each calibration curve was calculated. The absorbance of mixtures containing different ratios of TRMand XIP were measured at λ max 367.8nm corresponding to theconcentrationofTRMaloneandat λ 225 . 2 nm(Aiso)correspondingto total concentration of TRM and XIP in the mixture. Both thetotal concentration of TRM and XIP in the mixture and theconcentration of TRM alone were calculated using respectiveregression equations. Then, XIP concentration in the mixturewas calculated by subtracting TRM concentration from the totalconcentration. Ratiosubtractionspectrophotometricmethod Different aliquots of XIP containing 20–100 µ g were accuratelytransferred into a set of 10-mL volumetric flasks from its workingsolution and the volume was completed with methanol. Thespectra were divided by the spectrum of 10 µ g mL − 1 of TRM (as adivisor). The absorbencies in the plateau at λ above 367 nm weresubtracted from the corresponding spectra of the mixture andthentheproducedspectraweremultipliedbythespectrumofthedivisor. XIP concentration was determined at its λ max 230.8nmusing its corresponding regression equation while TRM wasmeasured at its λ max 367.8nm in the zero-order spectrum of themixture and its concentration determined from its correspondingregression equation. Chemometricmethods Constructionofthetrainingset Multilevel multifactor design was used for the construction of the calibration and validation sets. [25] A five-level, five-factorcalibration design was used. Different mixtures of TRM and XIP indifferentratioswereprepared(Table 1).Theabsorption spectraof the prepared mixtures were recorded and transferred to Matlab ® for subsequent data manipulation.Fifteen mixtures were used for building the calibration model,whiletenmixtureswerechosentobeusedasanexternalvalidationset. Several multivariate calibration models (CLS, PCR, and PLS)were constructed using the data obtained. Initial developedmodels were found to have high spectral residuals in the regionbelow230andabove285 nm;asaresult,thisregionwasrejected.For CLSmethod, construct CLS model with non-zero intercept. SelectionoftheoptimumnumberoffactorstobuildthePCRandPLSmodels Thecrossvalidationmethodwasused,leavingoutonesampleatatime,toselecttheoptimumnumberoffactors. [26] Givenasetof15calibrationsamples,thePCRandPLScalibrationswereperformed,andusingthiscalibration,theconcentrationofthesampleleftout www.drugtestinganalysis.com Copyright c 2010 John Wiley & Sons, Ltd. DrugTest.Analysis 2010 , 2 , 113–121 1 1 5 Spectrophotometric and spectrodensitometric determination of triamterene and xipamide Drug Testingand Analysis Table1. The concentrationof mixturesof xipamideandtriamtereneused in the training and validation sets Mixture No. Xipamide ( m g mL- 1 ) Triamterene ( m g mL -1 )1 442 43 34 35 56 67 58 49 610 6 611 6 212 2 513 5 214 2 415 4 516 5 517 5 318 3 219 2 320 3 421 4 222 2 223 2 624 6 325 3 664335654 - The concentrations of mixtures used in the validation set arehighlighted. waspredicted.Thepredictedconcentrationswerethencomparedwith the actual concentrations and the root mean square errorof cross validation (RMSECV) was calculated. The RMSECV wascalculated in the same manner each time a new factor was addedto the model. The maximum number of factors used to calculatethe optimum RMSECV was selected to be 8 (half the numberof samples + 1). [27] Visual inspection was used for selecting theoptimumnumberoffactors.Uponbuildingthemodelsautoscale,the data gave better results for both PCRand PLS. Constructionofthevalidationset Ten different mixtures of TRM and XIP were prepared by transfer-ring different volumes of their working standard solutions theirconcentrations given in Table 1. The developed models were ap-pliedtopredicttheconcentrationofTRMandXIPineachmixture. Spectrodensitometricmethod Intoasetof10-mLvolumetricflasks,differentaliquotsofTRM andXIP were accurately transferred from their working solutions; thevolumewasthencompletedwithmethanol.10 µ Lofeachsolutionwas spotted as bands of 6 mm width on TLC plates (20 × 10 cmwith 250 µ m thickness) using a Camag Linomat IV applicator. The bands were applied at 5 mm intervals and 10 mm from thebottom and sides. Linear ascending chromatogram developingto a distance of 8 cm was performed in a chromatographic tank previously saturated for 1 h with the developing mobile phaseconsisted of chloroform–methanol–ammonia solution (8:2:0.2,by volume) at room temperature. The peak areas were recordedusing scanning wavelength at 254 nm and the calibration curveswere constructed by plotting the integrated peak area versus thecorresponding concentrations of each drug and the regressionequations were computed. Applicationtopharmaceuticalformulation(Epitens ® tablets) The contents of 20 Epitens ® tablets were powdered and mixedwell. An accurately weighed portion of the powdered tabletequivalent to 75 mg of TRM and 15 mg of XIP was transferredinto 100-mL volumetric flask; 75 mL methanol was added andsonicated for 30 min, filtered, and then completed to volumewith methanol. The solution was diluted to obtain 100 µ g mL − 1 working solution using methanol as a solvent. The procedureof each method was followed and the concentration of TRMand XIP was calculated from the corresponding regressionequation. ResultsandDiscussion The main task of this work was to develop simple, sensitive andaccurateanalyticalmethodsforthedeterminationofTRMandXIPin their binary mixture either in bulk powder and pharmaceuticalformulationwithsatisfactoryprecisionforgoodanalyticalpractice(GAP). Isoabsorptivepointspectrophotometricmethod In this work, the so-called ‘isoabsorptive spectrophotometry’,developed by Erram and Tipnis, [28–30] is applied for simultaneousdetermination of TRM and XIP in their binarymixture. The theory of this method could be confirmed experimentallybyrecordingtheabsorbancespectraof8 µ g mL − 1 ofTRMandXIPseparately, and that of a mixture containing equal concentrationof TRM and XIP (4 µ g mL − 1 of each of TRM and XIP), as shown inFigure 2.In Figure 2, it can be seen that the mixture and the pure drugshave different absorbance spectra; meanwhile they possess thesame absorbance at their isoabsorptive point. These data permitsustoconclude,accordingtotheisoabsorptivespectrophotometrytheory, that the mixture ofboth drugs acts as a singlecomponentand gives the same absorbance value as pure drugs at theirisoabsorptive points. Thus, by measuring the absorbance valueat the chosen isoabsorptive point, the total concentration of themixture could be calculated as explained by the theory. [28–30] Byapplying the suggested procedure the absorbance at λ 225 . 2 nm(Aiso) for XIP was obtained over different ranges, while theconcentration of TRM in TRM and XIP mixture could be calculatedat its λ max 367.8nm without any interference from XIP. Thus theconcentration of XIP could be calculated by subtraction.Linear correlations were obtained between absorbance at367.8nm for TRM and its concentration in the range of 2–12 µ g mL − 1 and at 225.2nm for XIP and its concentration inthe range of 2–10 µ g mL − 1 from which the regression equationswere calculated and found to be:-For TRM at 367.8nm, A 1 = 0 . 0782C 1 + 0 . 0051 r 1 = 0 . 9998For XIP at 225.2 nm, A 2 = 0 . 1175C 2 + 0 . 0282 r 2 = 0 . 9999Where A 1 and A 2 are the absorbance of TRM and XIP,respectively, C 1 and C 2 are the concentration of TRM and XIP in DrugTest.Analysis 2010 , 2 , 113–121 Copyright c 2010 John Wiley & Sons, Ltd. www.drugtestinganalysis.com 1 1 6 Drug Testingand Analysis N. E. Wagieh etal . Figure2. Zeroorderabsorptionspectraof8 µ g mL − 1 ofTriamterene( ),8 µ g mL − 1 ofXipamide(- - - - )anda(1:1)mixturecontaining4 µ g mL − 1 of each of Triamterene and Xipamide ( . . . . ) using methanolas a blank. Table2. Determination of triamterene and xipamide in laboratory prepared mixtures by the proposed isoabsorptive and ratio subtractionspectrophotometricmethodsRecovery % ∗ XIP byMixture ratio TRM:XIP Concentration ( µ g mL − 1 ) TRM at 367.8 nm Isoabsorptiveat 225.2 nm Ratio subtraction at 230.8 nm3:1 ∗∗ 9:3 100.1 102.0 100.02:1 8:4 99.5 98.5 99.01:3 2:6 99.5 101.5 101.51:2 4:8 101.5 100.0 100.31:1 5:5 100.4 100.2 100.34:1 8:2 100.5 99.5 99.8Mean ± SD 100 . 3 ± 0 . 75 100 . 3 ± 1 . 29 100 . 2 ± 0 . 82 ∗ Average of 3 determinations. ∗∗ The ratio of pharmaceuticalformulation. µ g mL − 1 ,respectively,andr 1 andr 2 arethecorrelationcoefficientsof TRM and XIP, respectively.Results given in Table 2 show that the method is valid fordetermination of TRM and XIP in different laboratory preparedmixtures. Ratiosubtractionspectrophotometricmethod First,thelinearityofXIPwasdeterminedintheconcentrationrange2–10 µ g mL − 1 at 230.8nm in the zero order spectra. Differentdivisor concentrations (2, 5, 7 and 10 µ g mL − 1 ) were tried. Thedivisor concentration 10 µ g mL − 1 of TRM was found to be thebestregardingaccuracyandprecisionwhenthemethodwasusedfor calculation of XIP concentration in its laboratory preparedmixtures.Second,thespectrumofthemixtureofXIPandTRMinmethanolwas divided by the spectrum of the divisor (10 µ g mL − 1 of TRM). The value of the absorbance in the plateau region at λ above367 nmwassubtractedfromthespectrumofthedividedmixture;the obtained spectrum was then multiplied by the spectrum of the divisor as shown in Figure 3. Finally, XIP concentration wasmeasuredfromthelastspectrumobtainedat230.8nm,whileTRMconcentration was determined from zero-order spectrum at its λ max 367.8nm.Linear correlation was obtained between absorbance of XIPat 230.8nm and its concentration in the range 2–10 µ g mL − 1 from which the regression equation was calculated and found tobe:For XIP at 230.8 nm, A = 0 . 1281C + 0 . 0573 r = 0 . 9998whereAistheabsorbance,Cistheconcentrationin µ g mL − 1 ,andr is the correlation coefficient. The proposed method was successfully applied for thedetermination of XIP in laboratory prepared mixtures containingdifferent ratios of XIP and TRM as given in Table 2. www.drugtestinganalysis.com Copyright c 2010 John Wiley & Sons, Ltd. DrugTest.Analysis 2010 , 2 , 113–121 1 1 7 Spectrophotometric and spectrodensitometric determination of triamterene and xipamide Drug Testingand Analysis Figure3. Absorption spectra of Xipamide and Triamterene mixture ( ), mixture after division by the divisor (- - - - ), mixture after subtraction of the constant( ) and mixture after multiplication by the divisor ( . . . . ) using methanolas a blank. Table3. Results ofdetermination of xipamide and triamterene in the validation set using the proposedmultivariate methodRecovery % ∗ Xipamide TriamtereneMixture No. CLS PCR PLS CLS PCR PLS1 99.5 100.0 100.0 99.0 98.3 98.32 99.3 99.7 99.7 100.0 99.7 99.73 100.4 100.2 100.2 99.2 99.8 99.84 100.3 100.5 100.5 98.7 99.8 99.85 100.8 101.3 101.3 98.8 99.6 99.66 101.8 99.8 99.8 97.4 97.6 97.67 100.2 100.0 100.0 98.3 99.0 99.08 100.3 99.7 99.7 103.0 100.5 100.59 101.3 100.7 100.7 98.5 99.8 99.810 100.5 100.5 100.5 101.0 99.5 99.5Mean ± SD 100 . 4 ± 0 . 75 100 . 2 ± 0 . 51 100 . 2 ± 0 . 51 99 . 4 ± 1 . 60 99 . 4 ± 0 . 84 99 . 4 ± 0 . 84 ∗ Average of 3 determinations. Chemometricmethods In this method, different chemometric approaches were appliedforthedeterminationofXIPandTRM,includingCLS,PCRandPLS. These multivariate calibrations were useful in spectral analysisbecausethesimultaneousinclusionofmanyspectralwavelengthsinsteadofasinglewavelengthgreatlyimprovedtheprecisionandpredictive ability. [31] The first step in the simultaneous determination of the com-ponents by multivariate calibration methods involves construct-ing the calibration matrix for binary mixture. The calibrationset was obtained by using the absorption spectra of a setof 15 mixtures of XIP and TRM with different ratios of eachcomponent as given in Table 1. Better results were obtainedupon rejecting the spectral region above 285nm and below230 nm.In this study, the ‘leave one out’ cross validation method wasused and the RMSECV values of different developed models werecompared. Four factors were found suitable for both PCR andPLS models. To validate the prediction ability of the suggestedmodels, the validation set given in Table 1 was used to predictthe concentration of TRM and XIP, where satisfactory results wereobtained as shown in Table 3. The predicted concentrations of the validation samples wereplotted against the true concentration values. This was used todetermine whether the model accounted for the concentrationvariation in the validation set. All plots had a slope of nearly oneand an intercept close to zero. TheRMSEPwasanotherdiagnostictoolforexaminingtheerrorsinthepredictedconcentrations;itindicatesboththeprecisionandaccuracy. [26] The RMSEP values were 0.26, 0.20 and 0.21 for TRMand 0.21, 0.33 and 0.35 for XIP indicating the high predictiveabilities of the three models. DrugTest.Analysis 2010 , 2 , 113–121 Copyright c 2010 John Wiley & Sons, Ltd. www.drugtestinganalysis.com
