A high-performance liquid chromatographic method for benznidazole quantitation in plasma of patients with Chagas disease

Clin Chem Lab Med 2011;49(1):77–82 � 2011 by Walter de Gruyter • Berlin • New York. DOI 10.1515/CCLM.2011.014

2011/303

Article in press - uncorrected proof

A high-performance liquid chromatographic method

for benznidazole quantitation in plasma of patients

with Chagas disease

Laura Guerrero1, Ma Jesus Pinazo2, ElizabethPosada2, Joaquim Gascon2, Josep Ribas1

and Dolors Soy1,*1 Pharmacy Service, Hospital Clınic Barcelona/IDIBAPS,University of Barcelona, Barcelona, Spain2 Tropical Care Section, Hospital Clınic Barcelona/IDIBAPS, University of Barcelona, Barcelona Centre forInternational Health Research (CRESIB), CIBEREpidemiologıa y Salud Publica (CIBERESP), Barcelona,Spain

Abstract

Background: Chagas disease is endemic in Latin America,affecting 16–18 million people with more than 100 millionexposed to risk of infection. Its etiological agent is Trypa-nosoma cruzi. To date, benznidazole is the only treatment ofChagas disease available in Europe.Methods: A high-performance reversed-phase isocratic liq-uid chromatographic method for benznidazole analysis inhuman plasma is described. The mobile phase consists of60% ultrafiltered water and 40% acetonitrile. Samples wereprecipitated with trichloroacetic acid (0.3 M) (1/1, v/v). Theinjection volume was 100 mL. Benzocaine was used as inter-nal standard.Results: The assay was linear over a benznidazole concen-tration range of 1.6–100 mg/mL. The method showed goodagreement of results (ns15): inaccuracy (5.6%), intra- andinter-day variability (1.1% and 3.9%, respectively), recovery(94.9%), limit of detection (0.8 mg/mL), lower limit of quan-titation (1.6 mg/mL) and acceptable stability over 24 h inthe auto-sampler. Only 25 samples (58%) showed valueswithin the therapeutic range. Three samples were subthera-peutic and 15 were in the toxic range.Conclusions: The method offers a fast and simple approachto determining benznidazole in human plasma which couldbe of use in pharmacokinetic and safety studies.

*Corresponding author: Dolors Soy, PharmD, PhD, PharmacyService, Hospital Clınic de Barcelona, C/Villarroel, 170,08036 Barcelona, SpainPhone: q34 93 2275479, Fax: q34 93 2275457,E-mail: [email protected] May 26, 2010; accepted July 21, 2010;previously published online November 18, 2010

Keywords: analytics; benznidazole; Chagas disease; highpressure liquid chromatography; human plasma; validation.

Introduction

Chagas disease is endemic in Latin America, particularly inpoor rural areas of Mexico, Central America and SouthAmerica. It affects approximately 16–18 million people, andmore than 100 million people are exposed to the risk ofinfection (1). Its etiological agent is Trypanosoma cruzi, ahemoflagellate protozoan (family Trypanosomatidae, orderKinetoplastida), whose life cycle involves obligatory passagethrough vertebrate hosts (mammals, including humans) andinvertebrate hosts (hematophagous triatomine bugs) in aseries of stages (2). However, due to other possible routesof infection, such as mother to child, blood transfusion andorgan transplantation, together with the increased populationmovement, the disease and the transmission of T. cruzi areno longer limited to Latin America, and now has become aproblem worldwide (3–5).

The symptoms of Chagas disease vary over the course ofthe infection. In the acute stage, symptoms are usually mildand produce no more than local swelling at the site of infec-tion or unspecific symptoms. As the disease progresses overthe course of many years, serious chronic symptoms canappear, such as cardiomyopathy and/or megasyndromes inthe digestive tract. If untreated, chronic cardiac disease isoften fatal. At present, two compounds are available fortreatment of Chagas disease: benznidazole and nifurtimox,although benznidazole is only available in Europe (6).

To date, limited data is available on the pharmacokineticsof benznidazole. In order to optimize dosage, it would bevery helpful to establish its therapeutic range. Previous invitro studies have determined not only the minimal trypa-nosomicidal concentration of benznidazole, 3–6 mg/mL, butalso the toxic concentrations. Concentrations of benznidazoleG20 mg/mL seem to be related to a higher-risk of toxicity,mainly dermatological reactions (7–9).

High-performance liquid chromatography (HPLC) can beconsidered as a suitable technique for quantitation of benz-nidazole in human plasma. The sensitivity and precision ofHPLC and its applicability to a wide variety of compoundshas led to its use in clinical laboratories for monitoring of avariety of therapeutic agents, and also in pharmacokinetic,safety and metabolism studies (10, 11).

The goal of this study was to develop a new assay toquantify benznidazole in human plasma for further imple-mentation in pharmacokinetic and safety studies.

Brought to you by | Glasgow University LibraryAuthenticated | 130.209.6.50

Download Date | 5/14/13 2:29 AM

78 Guerrero et al.: HPLC method for benznidazole quantitation in human plasma


Materials and methods

Reagents

Benznidazole was obtained from F. Hoffmann-La Roche� Ltd. (SaoPaulo, Brazil). Trichloroacetic acid (TCA) was purchased from Flu-ka Chemie AG (Buchs, Switzerland). Acetonitrile gradient 240/far-ultraviolet (UV)-HPLC grade was acquired from Sharlau, S.A.(Barcelona, Spain). Dimethyl sulfoxide (DMSO) and benzocaine,the internal standard (IS), were obtained from Fagron Iberica S.A.U.(Terrassa, Spain).

Instrumentation

HPLC was performed using an Agilent 1100 liquid chromatographysystem (Agilent Technologies Spain, S.L., Madrid, Spain). Com-ponents consisted of a vacuum degasser, a gradient pump, an autosampler and a variable wavelength UV detector. Data acquisitionwas performed using the appropriate software for signal treatment(Agilent Chemstation Rev.B.03.01).

High-performance liquid chromatography

The HPLC method used in this study was based on the methodproposed by Morilla et al. for determining liposomal benznidazole(9). The stationary phase was a Kromasil 100-5 C18 silica basedcolumn (250=4.0 mm). The mobile phase consisted of 60% ultra-filtered water and 40% acetonitrile. The flow rate was fixed at0.9 mL/min and the injection volume at 100 mL. Absorbance wasmonitored at 324 nm.

Drug analysis and calibration

Calibration curves were constructed using drug free plasma. Theobserved peak-area ratio with reference to the IS was used to assessdrug concentrations. The benznidazole concentration in human plas-ma samples was calculated using the linear regression equation ofthe peak area ratio against the concentration ratio for the calibrationcurve.

(A /A )snqm=(wBNZ x/wBen x) w1xBNZ Ben theor theor

where ABNZ is the benznidazole area (mAu); ABen is the benzocainearea (mAu); wBNZtheorx is the benznidazole theoretical concentration(mg/mL); wBentheorx is the benzocaine theoretical concentration(mg/mL); m is the slope; and n is the ordinate where the BNZconcentration equals 0.

Sample collection

Samples were obtained from patients treated for Chagas infectionwith benznidazole (300 mg/day). These patients were enrolled in aclinical study designed to evaluate benznidazole adverse effects.This clinical study was approved by the Institutional Review Boardat Hospital Clinic (Barcelona). All plasma samples were centrifugedat 5000=g for 10 min. Plasma was separated and stored at –408Cuntil analysis.

Preparation of standard and quality controls

Stock standard and quality control solutions Two mg of benz-nidazole was weighed on an analytical balance (LE3245S, Sarto-rious Mechatronics Spain, S.A.U, Madrid, Spain) and 1 mL ofDMSO was added. This solution was diluted (1/4) with ultrapure

water to obtain stock solution A (500 mg/mL). Benzocaine (IS)stock solution was prepared by adding 10 mg of substance to 1 mLof DMSO. This solution was diluted (1/10) with ultrapure water toobtain stock solution B (1 mg/mL). Both solutions were preparedfreshly just before analysis.

Stock solution A containing 500 mg/mL of benznidazole wasused to prepare both standard and quality control material. Fromstock solution A, three levels of quality control (working standardsolutions) containing 3.1 mg/mL, 12.5 mg/mL and 100 mg/mL ofbenznidazole were prepared by diluting into drug-free plasma. Thesedilutions were also spiked with benzocaine (IS) (stock solution B)to achieve a final concentration of 300 mg/mL.

Standard preparation Plasma control specimens were spikedwith the appropriate volume of stock solution A to attain severalstandards with different benznidazole concentrations, and with45 mL of (IS) stock solution B. Finally, 75 mL of drug-free plasmawas added to obtain a final volume of 150 mL for the standards.These standards were mixed, and afterwards the mixture was pre-cipitated with 150 mL of TCA (0.3 M), (1/1, v/v). Next, they werecentrifuged at 8000=g for 10 min. Finally, 100 mL of the super-natant was injected into the chromatographic system.

Sample preparation All human plasma samples were spikedwith 45 mL of (IS) stock solution B to a final concentration of300 mg/L of benzocaine. This mixture was precipitated with150 mL of TCA 0.3 M (1/1, v/v). Afterwards, they were centrifugedat 8000=g for 10 min and 100 mL of the supernatant was injectedinto the chromatographic system.

Validation criteria

In accordance with the US Food and Drug Administration’s ‘‘Guid-ance for Industry, Bioanalytical Method Validation’’, and the Euro-pean Agency for the Evaluation of Medicinal Products Guidelineon Validation of Analytical Procedures, the following seven criteriawere evaluated: linearity, accuracy, precision, recovery, limit ofdetection (LOD), lower limit of quantitation (LLOQ) and stability(12–16). In addition, quality control samples were analyzed in par-allel to establish the suitability of the method.

Linearity For method validation, the range of linearity was estab-lished based on three separate assay runs of freshly prepared cali-bration standards. A seven-point standard curve of benznidazole wasconstructed by drawing the ratio between the peak area of benzni-dazole and the peak area of benzocaine versus the ratio betweenbenznidazole concentration and benzocaine concentration. The con-centration of benznidazole ranged from 1.6 to 100 mg/mL.

Linearity was assessed by determining coefficient of correlation(r2), Pearson’s coefficient (r), and the (tstudent) for the slope.

Accuracy and precision Accuracy of an analytical method isdefined as the agreement between results obtained by the analyticalmethod and the true value (12, 13). Precision describes the degreeof the similarity. The accuracy of the method was performed usinga recovery study with 15 drug-spiked samples at three concentra-tions (100, 12.5 and 3.1 mg/mL) measured in one run performed ina single day.

The same drug concentrations were analyzed over three consec-utive days to determine inter-day precision. Accuracy and precisionwere calculated as follows (Eqs. 2 and 3, respectively):



Guerrero et al.: HPLC method for benznidazole quantitation in human plasma 79


E (%)s wBNZx ywBNZ x /wBNZ x=100 w2xŽ .R exp theor theor

where ER is accuracy; wBNZtheorx is benznidazole theoretical con-centration (mg/mL); and wBNZexpx is benznidazole experimentalconcentration (mg/mL).

CV (%)s(SD/X )=100 w3xave

where CV is the coefficient of variation; SD is the standard devia-tion; and Xave is the average drug concentration (mg/mL).

For the analysis to be considered acceptable, no more than one-third of the quality control tests were allowed to deviate from thenominal concentration by more than 15%, and at least 50% of theresults from each quality control sample needed to be within 15%of the nominal concentration (15, 16).

Recovery Recovery (R) was determined by comparison of theanalytical results for extracted samples at three concentrations (low,medium, and high) with unextracted standards (100% recovery).Recovery was calculated as follows:

R (%)s(A A )/(A /A )=100 w4xBNZ Exp/ Ben Exp BNZ Cont Ben Cont

where ABNZ Exp is the benznidazole area of the spiked extract (mAu);ABen Exp is the benzocaine area of the spiked extract (mAu);ABNZ Cont is the benznidazole area of the unextracted control (mAu);and ABen Cont is the benzocaine area of the unextracted control(mAu).

Limit of detection (LOD) Several samples with known concen-trations of benznidazole (plus the IS) were prepared fresh. The LODwas established by visual evaluation of the minimum concentrationat which the benznidazole signal can be reliably detected.

Lower limit of quantitation (LLOQ) The LLOQ is the lowestamount of analyte in a sample that can be determined with a pre-cision and accuracy that is not -"20% (9). This parameter wasestablished by analysis of seven samples with known concentrationsof benznidazole (linearity assay). Bias (ER%) and precision (CV%)at the LLOQ were evaluated by a recovery study of five replicates.

Specificity Specificity was assessed in the presence of metroni-dazole (antimicrobial and antiparasitic drug) to guarantee that noresidual interfering peaks were present.

Stability

Freeze/thaw stability The stability of benznidazole was deter-mined after three freeze (–408C) and thaw cycles of three drug-spiked samples at concentrations of 100 mg/mL, 12.5 mg/mL, and3.1 mg/mL, all measured in one run performed in one day. Ben-zocaine was spiked into the samples after three freeze and thawcycles, just prior to analysis.

Post-preparative stability The stability of processed samples atroom temperature, including the time in the auto-sampler, was deter-mined using three aliquots at different concentrations. Resultsobtained from fresh samples were compared with those obtainedfrom the same samples after 24 h in the auto-sampler. In both casesstability was determined by comparison of the analytical results forextracted samples at three concentrations (low, medium, high), afterthree freeze (–408C) and thaw cycles, or after 24 h in the auto-sampler, with freshly extracted standards. Stability (S) was calcu-lated as follows:

S (%)s(A /A )/(A /A )=100 w5xBNZ/j Ben/j BNZ Cont Ben Cont

where ABNZ/j is the benznidazole area; j is the three freeze (–408C)and thaw cycles or after 24 h in the auto-sampler extract (mAu);ABen/j is the benzocaine area of the spiked extract (mAu) in theaforementioned scenarios (j); ABNZ Cont is the benznidazole area ofthe freshly extracted control (mAu); and ABen Cont is the benzocainearea of the freshly extracted control (mAu).

The non-parametric Wilcoxon test was applied to compare resultsand p-0.05 was used for statistical significance. Statistical analysiswas performed with SPSS 15.0 for Windows (SPSS Inc., Chicago,IL, USA).

Results

HPLC assay

Benznidazole and the IS were extracted quantitatively fromhuman plasma after precipitation. Under the experimentalconditions described in the previous section, benznidazoleand the IS were eluted within the retention windows of4–5.3 min and 7–8 min, respectively. Representative chro-matograms of drug-free human plasma and quality controlsamples are shown in Figure 1A and B. Metronidazoleshowed no interfering peaks in the analysis of benznidazole(metronidazole retention time: 2.9–3 min).

Assay validation

The correlation between drug concentration and peak areawas excellent for human plasma across the concentrationrange studied (1.6–100 mg/mL). Calibration curves (ns3),the correlation coefficient (r2) and the Pearson coefficient (r)were calculated. These were: Ys5.7309X–0.0255(r2s0.9975; rs0.9987); Ys5.2487X–0.0115 (r2s0.9961;rs0.9980) and Ys5.6618X–0.0152 (r2s0.9991;rs0.9995). X represents the concentration ratio and Y thepeak area ratio. The Pearson coefficient was not significantlydifferent from the line of unity (p)0.05) in any of the cases.Inaccuracy and precision were assessed by analysis of qualitycontrol samples. Results are shown in Table 1.

The LOD was 0.8 mg/mL. The LLOQ was 1.6 mg/mL(bias and precision: ER%s6.7% and CV%s5.6%, respec-tively). Typical chromatograms of the LLOQ are shown inFigure 2A and B.

The mean recovery remained fairly constant (99.6%,97.5% and 99.1%) at three different concentrations of benz-nidazole (100, 12.5 and 3.1 mg/mL, respectively). Resultsfrom the stability analysis revealed no loss of benznidazoleafter three freeze (–408C) and thaw cycles. Results areshown in Table 2. No significant differences were seen inbenznidazole and benzocaine chromatographic signals afterbenznidazole controls had been maintained for 24 h in theauto-sampler. S (%)s98.4% at 100 mg/mL; S (%)s98.6%at 12.5 mg/mL; S (%)s102.1% at 3.1 mg/mL.

Patient samples

Demographic data for all patients (ns12; 59% female) stud-ied showed a mean age of 37.6"9.5 years. Blood samples





Figure 1 Chromatogram of (A) blank human plasma and (B) spiked control benznidazole plasma (50 mg/mL).

Table 1 Results for validation parameters.

3.1 mg/mL 12.5 mg/mL 100 mg/mL

Intra-day inaccuracy –8.2 5.4 5.6(ER%) (ns5)Inter-day inaccuracy –2.3 10.8 8.4(ER%) (ns15)Intra-day precision 2.4 0.4 0.6(CV%) (ns15)Inter-day precision 3.5 3.9 4.2(CV%) (ns15)

were obtained just prior to starting treatment with benzni-dazole, and at trough concentrations on days q15, q30,q45 and q60. Forty-nine samples of a total of 60 that werescheduled to be collected were analyzed. In three patients,drug treatment was stopped before day q45 (ID �2 and ID�3) or at day q60 (ID �8). Therefore, no blood sampleswere available for these patients on these particular days.Furthermore, the sample prior benznidazole treatment wasmissing in six patients. In one patient, benznidazole wasstopped on day q14 due to adverse effects.

Benznidazole plasma concentrations showed values withinthe therapeutic range (3–6 mg/L) in only 58% of all samplesafter receiving the standard dose of 300 mg/day (male, 44%vs. female, 56%; ps0.28). Three (one male and two female)out of 43 samples (7%) were subtherapeutic (-3 mg/L),

while 15 samples out of 43 were clearly above the recom-mended therapeutic range ()6 mg/L) (male, 53% vs. female,47%; ps0.34). Additionally, no significant differences wereobserved between groups with respect to age (ps0.08).

Discussion

A specific HPLC method with suitable specificity, sensitivity,precision and accuracy has been developed to measure benz-nidazole in human plasma. This method allows rapid deter-mination of benznidazole concentrations in human plasma,offering a simple approach to sample preparation. Moreover,it is accurate, reproducible and specific for benznidazole.

In our study, the intra-day precision (five replicates at threedifferent concentrations) showed values -2.4% in all cases.The inter-day precision (five replicates analyzed on threeconsecutive days) was -4.2%. Moreover, inaccuracy was-5.6% and the LLOQ was set at 1.6 mg/mL. Specificity ofthe described analytical method was documented since noendogenous peaks co-eluted with benznidazole or benzo-caine. Additional analysis showed that metronidazole had adetection time that was different from benznidazole and ben-zocaine. Other nitroimidazole agents, such as dimetridazoleand misonidazole were not tested since they are not approvedfor human use.



Guerrero et al.: HPLC method for benznidazole quantitation in human plasma 81


Figure 2 Chromatogram of (A) lower limit of quantification (LLOQ), including magnification of the region of interest (B).

Table 2 Results from the stability analysis of benznidazole inhuman plasma after three freeze (–408C) and thaw cycles.

ABNZ/ABen ABNZ/ABen p-Valueat baseline after three cycles

100 mg/mL 1.91 1.95 )0.0512.5 mg/mL 0.24 0.24 )0.053.1 mg/mL 0.06 0.06 )0.05

ALNZ/AOflox, benznidazole area/benzocaine area; p-0.05 to besignificant.

To our knowledge, four studies on measurement of benz-nidazole in plasma have been published previously (8, 9, 17,18). Raaflaub et al. (8) analyzed benznidazole in blood plas-ma by differential pulse polarography (reduction of the nitrogroup in the range of –0.50 to –0.65 V vs. saturated calomelelectrode). They obtained an overall recovery (SD) of 78.2(3.9%). In another study, Morilla et al. (9) used an HPLCtechnique to quantify liposomal benznidazole in several bio-logical matrices. They obtained a recovery of 86% and inac-curacy was -6.4% across the entire range of concentrationsstudied. However, it is difficult to compare our results withthese studies, due to intrinsic differences in either the meth-odology or the matrix analyzed.

Our data are consistent with the results reported by Waltonand Workman (17), who determined benznidazole and itsamine metabolite in biological materials. They deproteinized

blood plasma by the addition of two volumes of methanolfollowed by centrifugation and evaporation of the superna-tant under vacuum. The residue was resuspended in runningbuffer. Their recovery and imprecision were 90% and 3.2%,respectively. In another study, Workman et al. (18) quantifiedbenznidazole in plasma, ultrafiltrate, urine and tissue homo-genates using reversed-phase HPLC. The LLOQ reported inthis study was 0.2 mg/mL, and imprecision was -2.5%(eight replicates at 10 mg/mL). In our study, intra-day pre-cision (five replicates at three different concentrations) was-2.4% for all concentrations studied, and inter-day precision(five replicates analyzed on three consecutive days) was-4.2%. In contrast to this study, our method showed aLLOQ of 1.56 mg/mL. However, this value should not haveany impact in terms of clinical significance since it is suf-ficient to quantify clinically significant blood concentrations()3 mg/mL).

It is worth noting that in our study, sample preparationwas conducted by means of protein precipitation with theaddition of TCA (0.3 M) to plasma, followed by centrifu-gation. This straightforward methodology was found to bequantitative, with a recovery (SD) of 98.7% (1.1%), whichis clearly superior to those reported in the aforementionedstudies by other authors.

A central reason for requesting plasma benznidazole con-centrations is to assess toxicity associated with use of thisdrug. The value of therapeutic drug monitoring in this cir-





cumstance depends on the strength of the relationshipbetween plasma benznidazole concentrations and toxicity.Although correlation between drug concentrations and effi-cacy is well-established, its influence on toxicity is less well-known. Efforts continue to better characterize this relation-ship, and therapeutic drug monitoring may be of benefit inthis issue.

In conclusion, our HPLC method allows for fast, easy,reliable and inexpensive analysis of benznidazole in humanplasma. Benznidazole analysis, performed using an isocraticmode, is completed within 10 min. This method conforms toU.S. Food and Drug Administration (FDA) and EuropeanMedicines Agency (EMEA) guidelines for bioanalyticalmethods, and may therefore, be helpful for further imple-menting pharmacokinetic and safety studies.

Acknowledgments

The authors would like to thank Dr. Jaime Altcheh for helping usto obtain the parent drug (benznidazole).

Conflict of interest statement

Authors’ conflict of interest disclosure: The authors stated thatthere are no conflicts of interest regarding the publication of thisarticle. Research funding played no role in the study design; in thecollection, analysis, and interpretation of data; in the writing of thereport; or in the decision to submit the report for publication.Research funding: Barcelona Centre for Internacional HealthResearch (CRESIB). CIBER Epidemiologıa y Salud Publica (CIBE-RESP). We thank Fundacion Mundo Sano Espana for support ofour research on Chagas Disease.Employment or leadership: None declared.Honorarium: None declared.

References

1. Garcia S, Ramos CO, Serna JF, Vilas-Boas F, Rodrigues MM,Campos-de-Carvalho AC, et al. Treatment with benznidazoleduring the chronic phase of experimental Chagas’diseasedecrease cardiac alterations. Antimicrob Agents Chemother 2005;49:1521–8.

2. Coura JR, de Castro SL. A critical review on Chagas diseasechemotherapy. Mem Inst Oswaldo Cruz 2002;97:3–24.

3. Munoz J, Coll O, Juncosa T, Verges M, del pino M, Fumado V,et al. Prevalence and vertical transmission of Trypanosoma cruziinfection among pregnant Latin American women attending twomaternity clinics in Barcelona, Spain. Clin Infect Dis 2009;48:1736–40.

4. Gascon J, Bern C, Pinazo MJ. Chagas disease in Spain, the Unit-ed States and other non-endemic countries. Acta Trop 2010;115:22–7.

5. Piron M, Verges M, Munoz J, Casamitjana N, Sanz S, MaymoRM, et al. Seroprevalence of Trypanosoma cruzi infection inat-risk blood donors in Catalonia, Spain. Transfusion 2008;48:1862–8.

6. de Lana M, Lopes LA, Martins HR, Bahia MT, Machado-de-Assis GF, Wendling AP, et al. Clinical and laboratory status ofpatients with chronic Chagas disease living in a vector-con-trolled area in Minas Gerais, Brazil, before and nine years afteraetiological treatment. Mem Inst Oswaldo Cruz 2009;104:1139–47.

7. Raaflaub J, Ziegler WH. Multiple-dose kinetics of the trypa-nosomicide benznidazole in man. Drug Res 1979;29:1611–4.

8. Raaflaub J. Multiple-dose kinetics of the trypanosomicide benz-nidazole in man. Drug Res 1980;30:2192–4.

9. Morilla MJ, Benavidez PE, Lopez MO, Romero EL. Liposomalbenznidazole: a high-performance liquid chromatographicdetermination for biodistribution studies. J Chromatogr Sci2003;41:405–9.

10. Jehl F, Gallion C, Monteil H. High-performance liquid chro-matography of antibiotics. J Chromatogr 1990;531:509–49.

11. Skoog D, Holler FJ, Nieman TA. Principios de Analisis Instru-mental, 5th ed. Madrid (Spain): The McGraw- Hill Company;2001. Meyer V. Practical high-performance liquid chromatog-raphy, 4th ed. Switzerland: John Wiley & Sons, Ltd; 2004.

12. Department of Health and Human Services Food and DrugAdministration (U.S). URL: a. http://www.fda.gov/cder/guid-ance/4252fnl.pdf. wAccessed January 12, 2010x, b. http://www.fda.gov/cder/guidance/cmc3.pdf. wAccessed January 12,2010x, c. http://www.fda.gov/cber/gdlns/methval.pdf. wAccessedJanuary 12, 2010x.

13. The European Agency for the Evaluation of Medicinal Products(Europe). URL: a. http://www.emea.europa.eu/htms/vet/vetgui-delines/quality.htm wAccessed January 12, 2010x, b. http://www.emea.europa.eu/pdfs/human/ich/038195en.pdf wAccessedJanuary 12, 2010x.

14. Lang JR, Bolton S. A comprehensive method validation stra-tegy for bioanalytical applications in the pharmaceutical indus-try. J Pharm Biomed Anal 1991;5:361–75.

15. Christodoulou EA, Samanidou VF, Papadoyannis IN. Validationof an HPLC-UV method according to the European UnionDecision 2002/657/EC for the simultaneous determination of10 quinolones in chicken muscle and egg yolk. J Chroma-togr 2007;857:246–55.

16. Samanidou VF, Nisyriou SA, Papadoyannis IN. Developmentand validation of an HPLC method for the determination ofpenicillin antibiotics residues in bovine muscle according to theEuropean Union Decision 2002/657/EC. J Sep Sci 2007;30:3193–01.

17. Walton MI, Workman P. Reversed-phase high-performance liq-uid chromatographic method for the simultaneous determina-tion of the 2-nitroimidazole benznidazole and its aminemetabolite in biological materials. J Chromatogr 1986;375:190–6.

18. Workman P, White RA, Walton MI, Owen LN, Twentyman PR.Preclinical pharmacokinetics of benznidazole. Br J Cancer1984;50:291–03.



Documents

A high-performance liquid chromatographic method for benznidazole quantitation in plasma of patients with Chagas disease