Revista Română de Medicină de Laborator Vol. 10, Nr. 1, Martie 2008

Bioanalytical method validation

Silvia Imre1, Laurian Vlase2, Daniela Lucia Muntean1

1. Department of Drugs Analysis, University of Medicine and Pharmacy from Targu-Mures, Faculty ofPharmacy, Gheorghe Marinescu 38, 540139 Targu-Mures, Romania

2. Department of Biopharmacy and Pharmacokinetics, University of Medicine and Pharmacy „IuliuHatieganu”, Faculty of Pharmacy, Victor Babes 41, 400012 Cluj-Napoca, Romania

Abstract

A syntetic discussion on bioanalytical methods validation is presented from the point of view of regula-tory documents, scientific articles and books. The validation parameters are described, together with an exampleof validation methodology applied in the case of chromatographic methods used in bioanalysis, taking in accountto the recent Food and Drug Administration (FDA) guidelines and documents of the International Conference onHarmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH).

Introduction

The reliability of analytical method is amatter of great importance in analysis and is, ofcourse, a prerequisite for correct interpretationof data.

As it is known, analytical method vali-dation is an experimental procedure whichdemonstrates that a specific analytical methodgenerates reliable, accurate and precise infor-mation about a sample. Bioanalytical methodsare used for the quantitation of drugs and theirmetabolites in biological matrices. Should wevalidate a bioanalytical method is not a questionanymore, validation proving the quality of theanalyst’s work, users of bioanalytical data getconfidence in the results and it is required bythe regulatory agencies. Before using a bioana-lytical method for quantitative determinationsof drugs and their metabolites, an applicant lab-oratory must first demonstrate that the envis-

aged method fulfills a number of performancecriteria after following a method validationprotocol.

Publications on bioanalytical methodsvalidation

The scientific literature about analyticaland bionalytical methods validation is reachand includes different categories: guidelines ofthe European and US committees, review arti-cles, books, documents published by interna-tional conferences or congresses etc.

Since the publications of the Europeanand US committees at the beginning of ’90years, many laboratories have started to re-design their processes by involving analysts andstatisticians, in order to define strategies thatwill allow the fulfillment of the regulatory re-quirements, while being practicable and scien-

13

Revista Română de Medicină de Laborator Vol. 10, Nr. 1, Martie 2008

tifically consistent. The Romanian NationalAgency of Drugs harmonized their require-ments under the latest international regulationswhich provide assistance in developing bioana-lytical method validation used in human clinicalpharmacology, bioavailability and bioequiva-lence studies requiring pharmacokinetic evalua-tion.

Frank T. Peters and Hans H. Maurer19

made in 2002 an excelent summary of the mostimportant documents published since 1991:

• The review on validation of bioanalyti-cal methods published by Karnes et al. (1991) -intended to provide guidance for bioanalyticalchemists15.

• The Shah et al. report (1992) on theconference on "Analytical Methods Validation:Bioavailability, Bioequivalence and Pharma-cokinetic Studies" held in Washington in 199022

- guidance for bioanalysts for the next years;contains the parameters of bioanalytical meth-ods which should be evaluated, and some ac-ceptance criteria were established but no specif-ic recommendations on practical issues like ex-perimental designs or statistical evaluation hadbeen made.

• Hartmann et al. (1994) analyzed the1990 Conference Report performing statisticalexperiments on the established acceptance crite-ria for accuracy and precision9. Based on theirresults they questioned the suitability of thesecriteria for practical application.

• Hartmann et al. (1998) review on vali-dation of bioanalytical chromatographic meth-ods - theoretical and practical issues were dis-cussed in detail10.

• The Shah et al. (2000) report on an up-date conference of the 1990 Washington con-ference23 - template for the guidelines (2001) ofthe U.S. Food and Drug Administration(FDA)25.

• The documents of the InternationalConference on Harmonisation of Technical Re-quirements for Registration of Pharmaceuticalsfor Human Use (ICH) and approved by the reg-

ulatory agencies of the European Union, theUnited States of America and Japan: the firstdocument, approved in 1994, concentrated onthe theoretical background and definitions invalidation13, the second, approved in 1996, onmethodology and practical issues14. The recentbook edited by Ermer J. and Miller J.H. (2005)7

is the latest document about these topics. De-spite the fact, that these were focussed on ana-lytical methods for pharmaceutical productsrather than bioanalysis, they still contain helpfulguidance on some principal questions and defi-nitions in the field of analytical method valida-tion.

Compliance with the 2001 FDA guid-ance can be considered today a minimum re-quirement to test the performance of a bioana-lytical method. At the beginning of this docu-ment the FDA states very clearly that its guid-ance for bioanalytical method validation repre-sents its current thinking on this topic and thatan alternative approach may be used if such anapproach satisfies the requirements of applica-ble statutes and regulations. This statement al-lows bioanalytical laboratories to adjust ormodify the FDA recommendations, dependingon the specific type of bioanalytical methodused.

In addition to such important docu-ments, different scientific journal publishedtheir opinion on these aspects. Journals likeJournal of Chromatography B17 or ClinicalChemistry have established their own criteriafor validation. Other perspectives are includedin a recent valuable book edited by Chan C.C.et al. (2004)3.

It is also necessary to present the guidepublished in 1997 by La Société Française desSciences et Techniques Pharmaceutiques (SF-STP)5 that provided the bioanalyst, on the onehand, with a better understanding on the way toproceed and on the other hand, real data forqualifying his own computations that he couldperform using a commercial spreadsheet4, 6, 11. Itshould be noted that this guide was published

14

Revista Română de Medicină de Laborator Vol. 10, Nr. 1, Martie 2008

before the recent FDA’s guide and introducesnew concepts in three different areas: stages ofthe validation, test of acceptability of a methodand design of experiments to perform. Themain authors of SFSTP guide recently pub-lished (2003) an article which objectives wereto identify and explain the progress permittedby the SFSTP guide, point out some of the limi-tations and suggest ways to overcome them2.An interesting thing is that no references aboutthe recently published FDA guide was made inthis article, just the first FDA guide (1992) iscited, even if the FDA document published in2001 provides more detailed aspect regardingexperimental procedures.

Many other scientific articles try to adda practical point of view on bioanalytical andanalytical methods validation1, 8, 12, 16, 18, 20, 24, 26.

Current validation practice onbioanalytical methods validation

In today’s drug development environ-ment, highly sensitive and selective methodsare required to quantify drugs in matrices suchas blood, plasma, serum, or urine. Chromato-graphic methods are the most commonly usedtechnology for the bioanalysis of smallmolecules and the general terms presented be-low take in account to this type of analyticalmethod.

It is well accepted the FDA Guidancefor Industry, Bioanalytical Methods Validation(2001) as a reference for current validationpractice and a briefly description of it is givenhere.

a. Glossary

The general concepts could be ex-pressed as follows:

• Validation - the process of checking ifsomething satisfies a certain criterion.

• Analytical method - a comprehensivedescription of all procedures used in sample

analysis.• Analytical method validation - a proce-

dure employed to demonstrate that an analyticalmethod used for quantification of analytes in abiological matrix quantifies the analyte with adegree of accuracy and precision appropriate tothe task.

o Full validation: establishment ofall validation parameters to apply to sampleanalysis for the bioanalytical method foreach analyte.

o Partial validation: modification ofvalidated bioanalytical methods that do notnecessarily call for full revalidation.

o Cross-validation: comparison ofvalidation parameters of two bioanalyticalmethods.

In order to understand the validationprocess it is necessary to define the analyticalterms used, including the validation parameters(Figure 1):

• Specificity/selectivity – the ability ofthe method to measure and differenciate the an-alyte signal in the presence of components thatmay be expected to be present. There has beensome controversial discussion about the termi-nology for this validation characteristic. In con-trast to the ICH, most other analytical organisa-tions define this as selectivity, whereas speci-ficity is regarded in an absolute sense, as the“ultimate degree of selectivity” (IUPAC). Se-lectivity is the ability of the bioanalyticalmethod to measure and differentiate the ana-lytes in the presence of components that may beexpected to be present. Specificity is the abilityto assess unequivocally the analyte in the pres-ence of components that may be expected to bepresent. For example, in high-performance liq-uid chromatography with UV detection (HPLC-UV), a classic chromatographic method, themethod is specific if the assigned peak at a giv-en retention time belongs only to one chemicalentity; in liquid chromatography with massspectrometry detection (LC-MS) the detectorcould measure selective an analyte, even if this

15

Revista Română de Medicină de Laborator Vol. 10, Nr. 1, Martie 2008

is not fully separated from endogenous com-pounds etc. Despite this controversy, there is abroad agreement that specificity/selectivity isthe critical basis of each analytical procedure.

• Precision - the closeness of agreement(degree of scatter) between a series of measure-ments obtained from multiple sampling of thesame homogeneous sample under the pre-scribed conditions. Precision may be consideredat three levels: repeatability, intermediate preci-sion and reproducibility. As parameters, thestandard deviation, the relative standard devia-tion (coefficient of variation) should be calcu-lated for each level of precision.

o Repeatability expresses the analyti-cal variability under the same operatingconditions over a short interval of time(within-assay, intra-assay).

o Intermediate precision includes theinfluence of additional random effects with-in laboratories, according to the intendeduse of the procedure, for example, differentdays, analysts or equipment, etc. (between-

assay, inter-assay).o Reproducibility, i.e., the precision

between laboratories (collaborative or inter-laboratory studies), is not required for sub-mission, but can be taken into account forstandardisation of analytical procedures.• Accuracy: the degree of closeness of

the determined value to the nominal or knowntrue value under prescribed conditions. This issometimes termed trueness. It is expressed asbias% or relative error%.

• Robustness: a measure of its capacityto remain unaffected by small, but deliberatevariations in method parameters and providesan indication of its reliability during normal us-age

• Limit of detection (LOD): the lowestconcentration of an analyte that the bioanalyti-cal procedure can reliably differentiate frombackground noise.

• Lower limit of quantification (LLOQ):the lowest amount of an analyte in a sample thatcan be determined quantitatively with suitable

16

Fig. 1. Validation parameters of a bioanalytical method

Validation parameters

Specificity and

selectivity

Accuracy

Repeatability, intermediate precision and

reproducibility

LinearityLimit of

detection LOD

Lower limit of quantification

LLOQ and upper limit of quantification

ULOQ

Robustness

Analyte stability

Recovery

Dilution effect

Revista Română de Medicină de Laborator Vol. 10, Nr. 1, Martie 2008

precision and accuracy.• Upper limit of quantification (ULOQ):

the highest amount of an analyte in a samplethat can be determined quantitatively with pre-cision and accuracy.

• Standard curve: the relationship be-tween the experimental response value and theanalytical concentration (also called a calibra-tion curve); usually this relationship is linear.

• Quantification range: The range ofconcentration, including the LLOQ and ULOQthat can be reliably and reproducibly quantifi edwith suitable accuracy and precision throughthe use of a concentration response relationship.

• Recovery: the extraction efficiency ofan analytical process, reported as a percentageof the known amount of an analyte carriedthrough the sample extraction and processingsteps of the method.

• Dilution effect: The ability to dilutesamples originally above the upper limit of thestandard curve should be demonstrated by accu-racy and precision parameters in the validation.

• System suitability: determination of in-strument performance (e.g., sensitivity andchromatographic retention) by analysis of a ref-erence standard prior to running the analyticalbatch.

• Reinjection reproducibility: It is neces-sary to be determined if an analytical run has tobe reanalyzed in the case of instrument failure.

• Stability: the chemical stability of ananalyte in a given matrix under specific condi-tions for given time intervals.

Other definitions should be given:• Biological matrix: a discrete material of

biological origin that can be sampled and pro-cessed in a reproducible manner. Examples areblood, serum, plasma, urine, feces, saliva, spu-tum, and various discrete tissues.

• Stock solutions: the original solutionsprepared directly by weighing the referencestandard of the analyte and dissolving it in theappropriate solvent. Usually, stock solutions areprepared at a concentration of 1-3 mg/mL in

methanol or acetonitrile and kept refrigerated at−20ºC if there are no problems of stability orsolubility.

• Working solutions: solutions preparedfrom the stock solution through dilution in theappropriate solvent at the concentration request-ed for spiking the biological matrix.

• Calibration standard: a biological ma-trix to which a known amount of analyte hasbeen added or spiked. Calibration standards areused to construct calibration curves from whichthe concentrations of analytes in QCs and in un-known study samples are determined.

• Internal standard: test compound(s)(e.g., structurally similar analog, stable labeledcompound) added to both calibration standardsand samples at known and constant concentra-tion to facilitate quantification of the target ana-lyte(s).

• Sample: a generic term encompassing:o Blank: a sample of a biological ma-

trix to which no analytes have been addedthat is used to assess the specificity of thebioanalytical method.

o Quality control sample (QC): Aspiked sample used to monitor the perfor-mance of a bioanalytical method and to as-sess the integrity and validity of the resultsof the unknown samples analyzed in an in-dividual batch. They are also used to calcu-late the accuracy and precision of themethod.

o Unknown sample: a biologicalsample that is the subject of the analysis.

b. Validation methodology of thechromatographic methods applied for drugsdetermination in human plasma

Includes two phases:• pre-study method validation – method

developing; method validation; it is performedbefore the unknown samples analysis

• routine-run method validation – duringunknown samples analysis

17

Revista Română de Medicină de Laborator Vol. 10, Nr. 1, Martie 2008

Pre-study method validationSpecificityAs a first step of method validation,

specificity is verified using six different plasmablanks obtained from healthy human volunteerswho had not previously taken any medication.A general approach to prove the selectivity(specificity) of the method is to verify that: theresponse of interfering peaks at the retentiontime of the analyte is less than 20% of the re-sponse of an LLOQ standard, or the response atthe LLOQ concentration is at least five timesgreater than any interference in blanks at the re-tention time of the analyte; the responses of in-terfering peaks at the retention time of the inter-nal standard are ≤5% of the response of theconcentration of the internal standard used inthe studies.

Standard curve. Quantification rangeThe relationship between the detector

response and concentration should be demon-strated to be well defined and reproducible. Thecalibration curve model is determined usuallyby the least squares analysis. In general, a poly-nomial function is considered:

y = b0 + b1x + b2x2 + b3x3 + ...;for the linear model, the terms x2 and

larger are ignored and for the quadratic model,terms larger than x2 are not considered. Even ifthe use of the quadratic model is allowed andused extensively by some bioanalytical labora-tories, the use of linear regression modelsshould be attempted first. Usually, a deviationfrom the linear model should be investigatedand avoided. Nonlinearity could be due to in-jection techniques, sample holdup on glassware,cross-talk in MS/MS, interferences, and toowide a concentration range. Weighing functionsreduce the influence of values obtained forhigher concentrations on slope and intercept butselection of weighing should be justified. Fromstatistical considerations the most commonweighing function for LC-MS- and LC-MS/MS-based assays should be 1/x, due to thefact that variance in y increases in proportion to

the concentration. In HPLC-UV-based assaysthe use of 1/x2 weighed linear regression analy-sis can significantly reduce the LLOQ obtain-able, where the standard deviation of y varieswith x. A comparison between the weighedleast squares procedure and the conventionalleast squares calibration shows improvementsin accuracy at the lower end. The principal ad-vantage in this case is for clinical pharmacologyand pharmacokinetic studies when concentra-tion values being measured by the method arenear LLOQ.

Calibration is performed using singli-cate, duplicate or triplicate (it depends onmethod precision) calibration standards on fivedifferent occasions. The concentration rangeshould cover the expected concentration in bio-logical samples. A calibration curve shouldconsist of a blank sample (matrix sample pro-cessed without the IS), a zero standard (matrixsample processed with internal standard), and 6to 8 nonzero standards. The number of stan-dards can be increased for a complex curve or acurve covering a very large range. The simplestrelationship that provides acceptable backcalcu-lated concentrations for the standards should beused first to fit the calibration curve. If aweighting factor is used, it should be definedduring validation. Distribution of the residuals(% difference of the back-calculated concentra-tion from the nominal concentration) should beinvestigated. The calibration model is accepted,if the residuals were within ±20% at the lowerlimit of quantification (LLOQ) and within ±15% at all other calibration levels and at least2/3 of the standards met this criterion, includinghighest and lowest calibration levels. Calibra-tion standards not meeting the acceptance crite-ria should be eliminated from the calibrationcurve calculations.

Lower limit of quantificationThe lower limit of quantification is es-

tablished as the lowest calibration standard withan accuracy and precision less than 20%.

18

Revista Română de Medicină de Laborator Vol. 10, Nr. 1, Martie 2008

Accuracy and precisionThe within- and between-run precision

(expressed as coefficient of variation, CV%)and accuracy (expressed as relative differencebetween obtained and theoretical concentration,Bias%) of the assay procedure is determined byanalysis on the same day of five different sam-ples at each of the lower (2 or 3 x LLOQ),medium (30-50% of the ULOQ) and higher(80% of the ULOQ) levels of the consideredconcentration range and one different sample ofeach on five different occasions, respectively.Sometimes, the selected concentrations couldinclude values which are relevant in practice.

RecoveryThe recoveries at each of the previously

three levels of concentration and limit of quan-tification are measured by comparing the re-sponse of the treated plasma standards with theresponse of standards in solution with the sameconcentration of analytes as the prepared plas-ma sample.

StabilityStock solution stability: The stability of

the stock solutions of drug and internal stan-dards should be evaluated at room temperaturefor at least 6 hours. If the stock solutions arekept refrigerated or frozen over a period oftime, the stability over that period should beevaluated by comparing the response of theaged stock solution to that of a freshly preparedstock solution. Stock solution stability shouldbe performed at one concentration in at leastduplicate.

The stability of the analytes in humanplasma: it is investigated in four ways, in orderto characterize each operation during the pro-cess of bioequivalence studies: room-tempera-ture stability (RTS), post-preparative stability(PPS) in the autosampler, freeze-thaw stability(FTS) and long-term stability (LTS) (at a freez-ing temperature at which it is known that theanalyte is stable). For all stability studies, plas-ma standards at low and high concentrations areused. The acceptance criterion: the mean found

concentration should be within ±15% and CV%< 15%.

Four plasma standards at each of thetwo levels are prepared and let at room temper-ature four hours before processing (RTS study).After that the extracted samples are analyzedwith fresh standards.

Other four pairs are prepared, immedi-ately processed and stored in the HPLC au-tosampler (PPS study). The samples are inject-ed periodically over the expected longest stor-age times of the samples in autosampler beforeinjection. The extracted samples (ready to in-ject) kept at autosampler temperature is finallyanalyzed with fresh standards.

For the freeze-thaw stability (FTS),aliquots at the same low and high concentra-tions are prepared. These samples are subjectedto three cycles of freeze-thaw operations inthree consecutive days. After the third cycle thesamples are analyzed against calibration curveof the day. The mean concentration calculatedfor the samples subjected to the cycles and thenominal ones are compared.

For long-term stability (LTS), in thefirst validation day, there are injected and ana-lyzed four samples at each of low and high con-centrations, and values are calculated againstcalibration curve of the day. Other two sets withthe same plasma concentrations were stored infreezer and analyzed together with calibrationsamples after the expected storage period. Thevalues are calculated against calibration curveof the day and the mean values for the storedsamples and nominal concentrations are com-pared. The requirement for stable analytes isthat the difference between mean concentra-tions of the tested samples in various conditionsand nominal concentrations had to be in ±15%range.

Dilution studyThe ability to dilute samples with con-

centrations above the upper limit of quantifica-tion is also investigated. Plasma standards (n =5) with the concentration levels above the

19

Revista Română de Medicină de Laborator Vol. 10, Nr. 1, Martie 2008

ULOQ are diluted with blank plasma in order toget a concentration within the calibration range,then processed and analyzed, five samples inthe same run and one sample on five differentoccasions. The mean found concentration iscompared with the nominal value. The accuracyand precision had to be within ±15% range.

Routine-run method validationRequirements for the calibration curve

of the run:• Good fitting of the experimental data• 75% of the calibration points (including

LLOQ and ULOQ) has a bias less than 15%,except LLOQ when is acceptable a limit of20%

• Values falling outside can be discardedif this does not change the established model

The accuracy and precision of the vali-dated method is monitored to ensure that it con-tinued to perform satisfactorily during analysisof unknown plasma samples. To achieve thisobjective, a number of QC samples prepared induplicate at the three concentration levels (low-er, medium and higher) are analyzed in each as-say run together with the calibration standardsand unknown samples. At least 67% (four outof six) of the QC samples should be within 15%of their respective nominal values; 33% of theQC samples (not all replicates at the same con-centration) can be outside ±15% of the nominalvalue.

Conclusion

As a conclusion generally accepted byall those involved in this topic, any analyticalmethod validation should be performed consid-ering the facts underlined by Shah V.P. in 2006(21): „No conference report or guidance cancover „ALL ISSUES”, and/or „ALL WHATIFS”. No substitute for common sense. Each is-sue needs to be evaluated in full light of objec-tives and aims of analysis, scientific basis andproof for deviation or anomalous observation.

No substitute for Good Science. The WorkshopReport and/or the Guiance can provide only theguiding principles.”

Bibliography

1. Bansal S., DeStefano A. - Key Elements of Bi-oanalytical Method Validation for Small Molecules,The AAPS Journal 2007; 9 (1): E109-E114.2. Boulanger B., Chiap P., Dewe W. et al. - Ananalysis of the SFSTP guide on validation of chro-matographic bioanalytical methods: progresses andlimitations, J. Pharm. Biomed. Anal., 2003; 32(4-5),753-765.3. Chan C.C, Lam H., Lee Y.C., Zhang X.-M. –Analytical Method Validation and Instrument Per-formance Verification, Wiley-Interscience, 2004.4. Chapuzet E., N. Mercier, S. Bervoas-Martin, B.Boulanger et al. - Méthodes chromatographiques dedosage dans les milieux biologiques. Exemple d’ap-plication de la stratégie de validation. Raport d’unecommission SFSTP, S.T.P. Pharm. Pratiq. 1998;8(2): 81-/107.5. Chapuzet E., N. Mercier, S. Bervoas-Martin, B.Boulanger et al. – Méthodes chromatographiques dedosage dans les milieux biologiques: stratégie de va-lidation. Raport d’une commission SFSTP, S.T.P.Pharm. Pratiq. 1997; 7(3): 169-194.6. Chiap P., Ph. Hubert, B. Boulanger, J. Crom-men, Anal. Chim. Acta 1999; 391: 227-238.7. Ermer J., Miller J.H. – Method Validation inPharmaceutical Analysis, Wiley-VCH 2005.8. González A.G., Herrador M.A. - A practical gui-de to analytical method validation, including measu-rement uncertainty and accuracy profiles, TrACTrends in Anal. Chem. 2007; 26(3): 227-238.9. Hartmann C., Massart D.L., McDowall R.D. -An analysis of the Washington Conference Reporton bioanalytical method validation, J.Pharm.Biome-d.Anal. 1994; 12: 1337-1343.10. Hartmann C., Smeyers-Verbeke J., MassartD.L., McDowall R.D. - Validation of bioanalyticalchromatographic methods, J.Pharm.Biomed.Anal.1998; 17: 193-218.11. Hubert P., Ph. Chiap, J. Crommen, B. Boulan-ger et al. - Anal. Chim. Acta 1999; 391: 135-148.12. Hughes N.C., Wong E.Y.K., Fan J., Bajaj N. -Determination of Carryover and Contamination forMass Spectrometry – Based Chromatographic As-

20

Revista Română de Medicină de Laborator Vol. 10, Nr. 1, Martie 2008

says, AAPS Journal 2007; 9 (3): E353-E360.13. International Conference on Harmonization(ICH). Validation of Analytical Methods: Definiti-ons and Terminology. ICH Q2 A. 1994.14. International Conference on Harmonization(ICH). Validation of Analytical Methods: Methodo-logy. ICH Q2 B. 1996.15. Karnes H.T., Shiu G., Shah V.P. - Validation ofbioanalytical methods, Pharm.Res. 1991; 8: 421-426.16. Kelley M., DeSilva B. - Key Elements of Bioa-nalytical Method Validation for Macromolecules,The AAPS Journal 2007; 9 (2): E156-E163.17. Lindner W., Wainer I.W. - Requirements forinitial assay validation, J. Chromatogr. B 1998; 707:1-2.18. Nowatzke W., Woolf E. - Best Practices Du-ring Bioanalytical Method Validation for the Cha-racterization of Assay Reagents and the Evaluationof Analyte Stability in Assay Standards, QualityControls, and Study Samples, The AAPS Journal2007; 9 (2): E117-E122.19. Peters F.T., Maurer H.H. - Review: Bioanalyti-cal method validation – How, how much and why ?,www.gtfch.org/tk/tk68_3/Peters.pdf20. Rocci M.L., Jr. , Devanarayan V., HaugheyD.B., Jardieu P. - Confirmatory Reanalysis of Incur-red Bioanalytical Samples, The AAPS Journal 2007;9 (3): E336-E343.21. Shah V.P. - History of Bioanalytical Validation

and Regulation: Evolution of a Guidance Documenton Bioanalytical Method Validation, AAPS 3rd Bi-oanalyticalWorkshop on Quantitative BioanalyticalMethods Validation and Implementation: Best Prac-tices for Chromatographic and Ligand Binding As-says, CrystalCity, Arlington, VA., May1-3, 2006,www.aapspharmaceutica.com/meetings/files/64/Shah.pdf22. Shah V.P., Midha K.K., Dighe S. et al. - Analy-tical methods validation: bioavailability, bioequiva-lence and pharmacokinetic studies. Conference re-port, Pharm.Res. 1992; 9: 588-592.23. Shah V.P., Midha K.K., Findlay J.W. et al. -Bioanalytical method validation - a revisit with a de-cade of progress. Pharm.Res. 2000; 17: 1551-1557.24. Smolec J., DeSilva B, Smith W et al. - Bioana-lytical method validation for macromolecules in sup-port of pharmacokinetic studies. Pharm. Res. 2005;22(9): 1425-1431. 25. U.S.Department of Health and Human Servi-ces, Food and Drug Administration. Guidance forIndustry, Bioanalytical Method Validation.http://www.fda.gov/cder/guidance/4252fnl.pdf26. Viswanathan C.T., Bansal S., Booth B. et al -Quantitative bioanalytical methods validation andimplementation: best practices for chromatographicand ligand binding assays. Pharm. Res. 2007;24(10): 1962-1973.

21