Analyst

PAPER

Publ

ishe

d on

17

Dec

embe

r 20

12. D

ownl

oade

d by

Uni

vers

ity o

f So

uthe

rn C

alif

orni

a on

02/

04/2

014

17:3

5:13

.

View Article OnlineView Journal | View Issue

aDepartment of Pharmaceutical Chemistry

Sciences, Orissa, India. E-mail: msaquibhasbBioanalytical Division, Fortis Clinical Rese

Cite this: Analyst, 2013, 138, 1581

Received 16th November 2012Accepted 14th December 2012

DOI: 10.1039/c2an36701g

www.rsc.org/analyst

This journal is ª The Royal Society of

Development and validation of LC–MS/MS method forthe quantitation of lenalidomide in human plasmausing Box–Behnken experimental design

M. Saquib Hasnain,*a Shireen Rao,b Manoj Kr Singh,b Nitin Vig,b Amit Gupta,b

Abdulla Ansari,b Pradeep Sen,b Pankaj Joshib and Shaukat Ali Ansarib

For the determination of lenalidomide using carbamazepine as an internal standard, an ultra-fast stability

indicating LC–MS/MS method was developed, validated and optimized to support clinical advancement.

The samples were prepared by solid-phase extraction. The calibration range was 2–1000 ng mL�1, for

which a quadratic regression (1/x2) was best fitted. The method was validated and a 32 factorial was

employed using Box–Behnken experimental design for the validation of robustness. These designs have

three factors such as mobile phase composition (X1), flow rate (X2) and pH (X3) while peak area (Y1) and

retention time (Y2) were taken as response. This showed that little changes in mobile phase and flow

rate affect the response while pH has no affect. Lenalidomide and carbamazepine were stable in human

plasma after five freeze thaw cycles, at room temperature for 23.7 h, bench top stability for 6.4 h. This

method competes with all the regulatory requirements for selectivity, sensitivity, precision, accuracy, and

stability for the determination of lenalidomide in human plasma, as well as being highly sensitive and

effective for the pharmacokinetic and bioequivalence study of lenalidomide.

1 Introduction

B-cell malignancies of the plasma cell cause multiple myelomawhich is the second most common haematological malignancy.Lenalidomide is one of the novel compounds which is used forthe treatment of this disease, it is based on the molecularstructure of thalidomide and has an improved immunomodu-latory effect and safety prole.1,2 Lenalidomide (3-(40-amino-isoindoline-10-one)-1-piperidine-2,6-dione) is a new chemicalentity which is a synthetic derivative of glutamic acid and isstructurally close to thalidomide, it is used for the treatment ofpatients with a sub-type of myelodysplastic syndrome (MDS).3–8

Bioanalysis of biological matrices is usually carried out usingliquid chromatography–tandem mass spectrometry (LC–MS/MS). Sample preparation for bioanalysis is accomplished byusing methods such a s protein precipitation, liquid–liquidextraction or solid-phase extraction with robotic liquidhandling systems.9–11 Tohniya et al. developed and validated anLC–MS method for the determination of the thalidomideanalogue CC-5013 using umbelliferone as the internal stan-dard.12 But this method is inferior to the method developed byus with respect to retention time and recovery of the analyte andinternal standard. Hence our method can be used to obtainsuperior results. Bosch et al.in their critical reviews on the

, Seemanta Institute of Pharmaceutical

nain@gmail.com; Tel: +919310810422

arch Limited, Faridabad, Haryana, India

Chemistry 2013

analytical determination of thalidomide and its metabolitediscussed various analytical methods for the determination ofthalidomide.13 Almost all the described methods were achiraland based on HPLC using reverse phase separation with UV, MSor tandem MS detection. However, recently some newapproaches like chiral stationary phase-HPLC or electro-chromatography have been used. They showed that vancomycinchiral stationary phase-HPLC is a good choice for the separationand quantication of thalidomide enantiomers.13

A chromatographic system which was similar to thatdescribed in a previous article was used by Svensson et al. Apump that operated in constant pressure mode was connectedto an injector. This was equipped with a 2mLmobile phase loopfollowed by a sample injector having an internal loop volume of0.2 mL. A UV detector was connected to the columns using a 50mm i.d. fused silica tube. A Finnigan TSQ 700 triple quadrupolemass spectrometer equipped with an electrospray ion sourcewas used for performing LC–MS and LC–MS/MS analyses. Thespray emitter was used with a ow rate of 3–5 mL min�1 andthe spray voltage was 4.5 kV. The temperature was set at 220 �C.The mobile phase used was acetonitrile mixed with ammoniumacetate buffer 5 mM or acetic acid 0.1 M.14

Teo et al. describe a highly sensitive, rapid, selective andhigh throughput LC–MS/MS assay for TD in human plasma andsemen, with a calibration range of 2–250 ng mL�1. To preventspontaneous hydrolysis the matrices were rst stabilized with0.025 M Sorensen's citrate buffer at pH 1.5. It was stable whenstored at room temperature for 24 h and for up to three

Analyst, 2013, 138, 1581–1588 | 1581

Analyst Paper

Publ

ishe

d on

17

Dec

embe

r 20

12. D

ownl

oade

d by

Uni

vers

ity o

f So

uthe

rn C

alif

orni

a on

02/

04/2

014

17:3

5:13

. View Article Online

freeze–thaw cycles. Solid-phase extraction of the samples wasthen carried out. The extracted samples were analysed by LC–MS/MS using a reversed-phase column and an APCI interface inthe negative ion mode.15 Beg et al. used experimental designsoware for the optimisation of validation processes.16 Hencewe used Design-Expert 8.0 for the optimisation of the robust-ness of validation. For this we used two parameters peak areaand retention time.

The detector used mainly is a triple quadruple mass spec-trometer equipped with either an electron spray or an atmo-spheric pressure chemical ionization ion source. For thedetermination of lenalidomide in human plasma, a solid-phaseextraction method was developed with a lower limit of quanti-cation (LLOQ) of 2 ng mL�1, which was further validated. Thestability of these analytes in human plasma under the mostfavourable conditions was validated as part of the methodvalidation.

2 Materials and method2.1 Test compounds and internal standards

The drug lenalidomide was purchased from Vivan Life SciencesPvt Ltd., Mumbai (India), while carbamazepine was purchasedfrom Sigma Aldrich.

2.2 Biological matrix

This was purchased from Yash Laboratories, Shop no-9,Louiswadi, Thane (India). For calibration standards (CC) andquality control (QC) sample preparation, human plasma withK2EDTA was used and was chromatographically screened forinterfering substances. For this, blank plasma was run as part ofeach validation to overcome interference in the biologicalmatrix.

2.3 Reagents

Acetonitrile HPLC grade from J.T Baker, methanol HPLC grade,ammonium formate (AF), and Water HPLC grade were used.

2.4 Solutions

2 milimolar (2 mM) AF solutions were prepared appropriately.Mobile phase was prepared by adding 150 mL solution of 2 mMAF in 850 mL of acetonitrile. The diluent solution was 80 : 20acetonitrile and water. Washing solution was prepared bymixing, methanol and water in a ratio of 5 : 95.

2.5 Stock solutions

A 1 mg mL�1 strength stock solution of lenalidomide wasprepared in the diluent solution. This solution was used toprepare spiking solutions of the above working standard. Stockdilutions of 50 ng were prepared for spiking of calibrationstandard (CC) and quality control (QC) standards of lenalido-mide by using this stock solution. These samples (CC and QC)were used to evaluate the accuracy and precision of the methodand were also used for the determination of the LLOQ. Similarlystock solution of carbamazepine was used as an internal

1582 | Analyst, 2013, 138, 1581–1588

standard working solution. These stock solutions were screenedchromatographically and by MS for interfering elements.

2.6 LC–MS/MS conditions

The drug lenalidomide was determined using an AB SCIEX API-4000 triple quadruple mass spectrometer, a Schimadzu (Japan)HPLC instrument with an analytical column of Lichrocart 125-4.Lichrospher 60, RP-Select B with a particle size of 5 mm. Themobile phase used was a mixture of AF solution and acetonitrilein a ratio of 15 : 85, and the ow rate was kept at 1 mL min�1. A10 mL injection volume and a 3 min run time were set. Theretention time of lenalidomide and carbamazepine (IS) werereported to be 1.04 and 1.2 min respectively. The mass spec-trometer was used in the positive ion mode with turbo ion sprayionisation, and the following parameters were set for the massspectrometer, curtain gas on 12, nebuliser on 14, declusteringpotential on 41, focusing potential on 120, collision cell exitpotential on 12, ion spray voltage on 4000 V, temperature on450 �C, collision gas on 3 and entrance potential on 10 Vrespectively. The dwell time was 400 ms. Multiple reactionmonitoring (MRM) conditions were monitored for the analytesand IS. For lenalidomide it was 260.08/149.20, for carbamaze-pine 237.02/194.40. Data were analysed by using Analyst 1.5soware.

2.7 Calibration standards and quality control samples

For CC and QC samples, plasma was spiked with lenalidomidein the range of the 2–1000 ngmL�1. The QC samples and a set ofnine CC standards of pooled plasma were prepared and storedat �22 �C. In the validation process each run comprised of a setof CC and six replicates of QC samples at low (5.8), medium(414.7) and high (829.5) ng mL�1 levels respectively. The LLOQ(2 ng mL�1) were analysed to determine the accuracy andprecision of the method.

2.8 Method validation and experimental design

The validation of the method was evaluated by three analyticalruns. Each of them included nine CC with QC samples, two zerostandards (with IS) and a blank (without IS) also incorporated inthe runs. Two thirds of the CC standards were required to passindividually and the individual accuracy of the standard had tobe within �15% of nominal �20% of the LLOQ. Among thethree analytical runs at least one of the two CC standards at theLLOQ and the upper limit of quantication (ULOQ) level wererequired to pass and the CC coefficient of determinant (r2) hadto be greater than 0.98. Within-run and between-run accuracyhad to be within �15% at QC levels and individual accuracy ateach concentration had to be within �15%.

While method selectivity was determined by screening out 20blank plasma batches to check out the interference at theretention time of the analytes as well as that of the IS. Foracceptance of the method, it had to be less than 20% of themean peak response calculated from the analysis of LLOQ QCsamples at the expected retention time of the analytes and lessthan 5% of that of the IS.

This journal is ª The Royal Society of Chemistry 2013

Paper Analyst

Publ

ishe

d on

17

Dec

embe

r 20

12. D

ownl

oade

d by

Uni

vers

ity o

f So

uthe

rn C

alif

orni

a on

02/

04/2

014

17:3

5:13

. View Article Online

The stability of the method was checked by analysing the QCsamples of the analytes at lower and higher concentrations (n¼6), stored under various conditions. For acceptance, two third ofthe QC samples had to have individual accuracy within �15%andmean accuracy had to be�15%. The stability of the analytesin stored stock solution during long term storage for 43 dayswas determined by comparing the stored stock solution withfresh stock solution. The solution was deemed stable if themean difference between the two solutions was less than 5%.

The stability of solutions stored under different conditions,like room temperature, and refrigerator temperature for 6, 16and 43 days was assessed. The reference mix stability at 24 hand reinjection reproducibility were also determined.

Dilution integrity was performed to evaluate the accuratequantitation within the range of CC at a concentration greaterthan the ULOQ which can be diluted with the matrix. Theacceptance criterion for this was that the individual and meanaccuracy had to be less than �15% and individually two thirdsof the diluted QC samples had to meet the above criteria.

For the optimisation of the method and data analysis theBox–Behnken design (BBD) was applied using Design-Expert8.0.7.1 soware. BBD optimised the experiment using a 3n

factorial (where n ¼ 1, 2, 3.) having 3-independent variablesand 2-dependent variables as compared to other designs suchas, the central composite design and fractional factorial design.The 3-independent variables mobile phase (AF buf-fer : acetonitrile) composition from 10 : 90–20 : 80, ow ratefrom 0.8–1.2 and pH 5.5–6.5, and 2-dependent variables(response) peak area as well as retention time were run in thesoware and a total of 17 runs were obtained.

2.9 Extraction procedure

The required number of samples were withdrawn from thesample storage device and allowed to thaw at room tempera-ture. The thawed samples were vortex mixed to ensure completemixing of the contents. 50 mL aliquots of IS dilution (50 ngmL�1

of carbamazepine) were transferred into micro-centrifuge tubesand 300 mL of each sample were added followed by vortexmixing. 400 mL of water was then added and the mixture wasvortex mixed. The required number of Waters HLB cartridgeswere placed in a positive pressure solid-phase extractionassembly for sample preparation. The cartridges were condi-tioned with 1 mL of acetonitrile followed by 1 mL of water, thensamples were loaded into the cartridges and passed through thecartridges under constant pressure. The cartridges were washedwith 1 mL of washing solution followed by 1 mL of water and 1mL of acetonitrile elution solution under constant pressure,then dried in a nitrogen evaporator at 50 �C and 20 psi tocomplete dryness. The residue was reconstituted with 500 mL ofmobile phase and transferred into autosampler vials.

3 Results and discussion

The perfectness of the method was assessed on the basis oflinearity, sensitivity, selectivity, system suitability, stabilitystudy, precision and accuracy etc. On the basis of the results

This journal is ª The Royal Society of Chemistry 2013

obtained for the above parameters, the combination of mobilephase, ow rate and pH were selected for validation. Out ofseveral tried combinations, BBD suggested a combination of15 : 85 ratio of mobile phase (AF buffer : ACN), ow rate 1 mLmin�1 and pH 6 which resulted in a retention time of 1.04 forlenalidomide and 1.2 for the IS.

3.1 Sample extraction

We opted out various methods for sample extraction includingsolid-phase extraction, liquid–liquid extraction and proteinprecipitation etc. Among these solid-phase extraction was foundto be very useful due to the following two reasons: rst thematrix effect was reduced as a clean extract was obtained fromwhich we could obtain the appropriate sensitivity at the LLOQlevel i.e. 2 ng mL�1 and secondly the method was clinicallyapplicable which is necessary for pharmacokinetic studies. Amatrix effect was initially detected which was removed by usingthis method. Hence this method of extraction was considered tobe the best procedure.

Elution was done by using acetonitrile, yielding a very goodrecovery of 80.2% for lenalidomide and 94.4% for carbamaze-pine which indicates that this method is good in terms of therecovery achieved. For all the analytes and IS at the LLOQ level,peak shapes were symmetric, recovery was found to be consis-tent and highly sensitive.

3.2 Tandem mass spectrometry

The positive ion mode was used for product ion mass spectraand scanning of lenalidomide and carbamazepine. The scan-ned mass spectra of lenalidomide and carbamazepine gavesignals for the protonated molecular ions at m/z ¼ 260.08 andm/z ¼ 237.02 respectively and product ion mass spectra of theabove compounds were 149.20 and 194.40. Transition from m/z260.08 tom/z 149.20 andm/z 237.02 to 197.40 weremonitored inthe MRM mode for quantication.

3.3 Separation

In volunteers, differentmetabolites of lenalidomidewere presentand these metabolites may have caused problems in the separa-tion. To solve this problem, Lichrocart 125-4, Lichrospher 60, RP-Select B,with a 5mmparticle size and amobilephaseow rate of 1mL min�1 with a suitable post column splitter were used.

3.4 Sensitivity

Sensitivity was evaluated in one of the validation runs at theLLOQ (2 ng mL�1) level by processing six replicates of it forlenalidomide. The batch precision and accuracy at the LLOQlevel was 6.6% and 100.4% for lenalidomide using the IS ratiomethod. Chromatograms of lenalidomide and carbamazepineare shown in Fig. 1(A and B).

3.5 Matrix effect

Matrix effect is the suppression or enhancement of ionisation ofanalytes by the presence of matrix in the biological samples.Earlier it was found that the matrix effect was too high, that's

Analyst, 2013, 138, 1581–1588 | 1583

Fig. 1 Chromatograms at the LLOQ level (1A) for lenlidomide and (1B) forcarbamazepine.

Fig. 2 Blank chromatograms of (A) lenalidomide and (B) carbamazepine.

Fig. 3 Calibration curve for lenalidomide.

1584 | Analyst, 2013, 138, 1581–1588

Analyst Paper

Publ

ishe

d on

17

Dec

embe

r 20

12. D

ownl

oade

d by

Uni

vers

ity o

f So

uthe

rn C

alif

orni

a on

02/

04/2

014

17:3

5:13

. View Article Online

why the method was so designed that it reduces the matrixeffect signicantly. The % matrix factor at 2, 414.7 and 829.5was found to be 0.9, 1.0 and 1.0, respectively. The abovereported method showed that no matrix effect was found forplasma. Blank chromatograms of these components are shownin Fig. 2 above (Fig. 2A and B).

3.6 Selectivity

Selectivity was assessed by screening 20 different batches ofblank plasma which did not contain any interference plasmacomponents or other sources at the retention time of the ana-lytes and IS and it was shown that this method is highly selec-tive for lenalidomide.

3.7 Calibration curve regression

The calibration curve regression for lenalidomide was aquadratic regression (1/concentration). This gave the best tand coefficient of determination (r2) for validation and wasgreater than 0.998 which was in the acceptable range. Thecalibration curve is presented in Fig. 3. The average value for r2

was found to be 0.998.

3.8 Accuracy and precision

The accuracy (% nominal) for lenalidomide within batch andbetween batches were reported to be 99.2 and 100.4, respec-tively, while the % CV within batch and between batches werereported to be 6.3% and 6.6%, respectively, and are presented inTable 1.

Table 1 Precision and accuracy of lenalidomide in human plasma

Concentrationnominal/ng mL�1

Precision (%) Accuracy (%)

Withinbatch

Betweenbatch

Withinbatch

Betweenbatch

2.0 11.3 10.3 98.6 107.45.89 4.9 6.9 98.4 97.7414.7 5.4 5.9 99.2 98.4829.5 3.6 3.2 100.6 98.1

This journal is ª The Royal Society of Chemistry 2013

Table 2 Stability of lenalidomide in human plasma

Storage conditionsConcentration/ng mL�1

Accuracy(%)

Bench top stability (6.4 h) 5.89 98.5829.5 98.5

Bench top extraction stability 5.89 100.9829.5 99.2

Long term stability (43 days) 5.89 101.8829.5 102.6

Freeze–thaw stability (aer 5 cycle) 5.89 94.3829.5 97.9

In injector stability (50.15 h) 5.89 99.5829.5 101.9

Table 3 Summary of the design

Stdrun S. no

X1 AF buffercomposition(%)

X2 owrate/mLmin�1

X3

pH

Y1 peakarea/cm2

Y2retentiontime/min

16 1 15 1 6 697 998 1.047 2 10 1 6.5 663 925 1.085 3 10 1 5.5 657 635 1.0615 4 15 1 6 697 998 1.048 5 20 1 6.5 632 188 1.086 6 20 1 5.5 635 437 1.0712 7 15 1.2 6.5 642 286 1.0610 8 15 1.2 5.5 642 846 1.0811 9 15 0.8 6.5 661 075 1.069 10 15 0.8 5.5 653 474 1.072 11 20 0.8 6 623 547 1.061 12 10 0.8 6 635 438 1.0717 13 15 1 6 697 998 1.043 14 10 1.2 6 634 628 1.0714 15 15 1 6 697 998 1.0413 16 15 1 6 697 998 1.044 17 20 1.2 6 632 098 1.09

Paper Analyst

Publ

ishe

d on

17

Dec

embe

r 20

12. D

ownl

oade

d by

Uni

vers

ity o

f So

uthe

rn C

alif

orni

a on

02/

04/2

014

17:3

5:13

. View Article Online

3.9 Recovery

By comparing the peak area of the QC sample before extractionto the peak area aer extraction, recoveries were determined.Recoveries of lenalidomide at low, medium and high QC level

Table 4 ANOVA table for peak response

Parameters SS Df MS F-value

% AF 5.841 1 5.841 11.06Flow rate 5.873 1 5.873 1.11pH 1.271 1 1.271 0.24

Table 5 ANOVA table for retention time

Parameters SS Df MS F-value p

% AF 5.0 1 5.0 0.70 0Flow rate 2.0 1 2.0 2.80 0pH 0.0 1 0.0 0.00 1

This journal is ª The Royal Society of Chemistry 2013

were 76.5, 86.2 and 83.8%, respectively, while recovery of car-bamazepine was 94.4%. This shows that the method has verygood recovery of both analytes and IS.

3.10 Integrity of dilution

DIQC (dilution integrity quality control) samples were preparedand diluted 2 and 4 times with blank matrix, the % nominal ofDIQC samples at 2 and 4 times dilution were found to be 94.9and 94.2%, respectively, and the % CV were reported to be 2.9and 6.8%, respectively. The integrity of dilution was found to bewithin the accepted range.

3.11 Stability

To determine whether the analytes and IS were stable underdifferent storage and processing conditions, the duration of timerequired for sample processing was not too short to overcome theinstability issues.Hence the stability of the analytesduring sampletransport, storage and preparation were assessed using QCsamples (n¼ 6) at low,medium and high QC levels, by comparingthe mean value of the stability QC sample at each level with themean value of the same QC pooled freshly. It was found thatlenalidomidewas stable inhumanplasma. The results of differentstability studies for lenalidomide are summarized in Table 2.

From Table 2 it is clear that samples of the analytes and ISwere stable under the various stability conditions listed above.

3.12 Ruggedness

The ruggedness of the method was determined by differentanalysts. For the evaluation of ruggedness of the method, oneprecision and one accuracy batch were analyed using a differentcolumn of the same type by a different analyst using the same ordifferent instrument. The within batch accuracy (%nominal) forlenalidomide was 100.2% and the within batch precision (% CV)was 4.3%. The values reported above for lenalidomide indicatesthat themethodmay be used under slightly different conditions.

Robustness of test using experimental design. The capa-bilities to remain unaffected by small and deliberate changesin chromatographic conditions like mobile phase composi-tion, ow rate, column temperature etc. is known asrobustness. To test the robustness BBD was applied in which

p-value Model F-value Model p-value Prob > F

0.0127 25.75 0.0001 Signicant0.32670.6388

-value Model F-value Model p-value Prob > F

.4304 6.14 0.0129 Signicant

.1382

.000

Analyst, 2013, 138, 1581–1588 | 1585

Analyst Paper

Publ

ishe

d on

17

Dec

embe

r 20

12. D

ownl

oade

d by

Uni

vers

ity o

f So

uthe

rn C

alif

orni

a on

02/

04/2

014

17:3

5:13

. View Article Online

3-independent variables mobile phase (v/v% AF buffer)composition (X1), ow rate (X2) (mL min�1) and pH (X3) and2-dependent variables (response) peak area (cm2) (Y1) as wellas retention time (min) (Y2) were taken as robustnessparameters. 17 runs were obtained and each suggestedcombination were run on the system and results wereobtained (Table 3).

To minimise the effect of uncontrolled changes which maybias the response, experiments were designed in a randomisedmanner. The experimental design suggested second degree

Fig. 4 Effect of individual factors on response (Y1 and Y2). (A) for peak area (Y1) a

1586 | Analyst, 2013, 138, 1581–1588

quadratic equation with highest least square regression for Y1(r2 ¼ 0.9707) and Y2 (r

2 ¼ 0.9803) as compared to others.Themodelwasevaluated forpeakareaandretention timeonthe

basis of analysis of variance (ANOVA) to nd whether it wassignicant or not. Both responses showed that results were signif-icant. Equations for both responses were found to be as follows

Y1 (peak area) ¼ 6.980 � 10005 � 8544.50A � 2709.50C +

2340.25AB � 2384.75AC � 2040.25BC � 34597.13B2 �16104.62C2

nd (B) for retention time (Y2).

This journal is ª The Royal Society of Chemistry 2013

Paper Analyst

Publ

ishe

d on

17

Dec

embe

r 20

12. D

ownl

oade

d by

Uni

vers

ity o

f So

uthe

rn C

alif

orni

a on

02/

04/2

014

17:3

5:13

. View Article Online

Y2 (retention time)¼ 1.04 + 2.500� 10�003A + 5.000� 10�003B +

0.000C + 7.500 � 10�003AB � 2.500 � 10�003AC � 2.500 �10�003BC + 0.019A2 + 0.014*B2 + 0.014C2

Table 4 shows that the ANOVA results were satisfactory forresponse Y1 and a predicted model F-value of 25.75 and a modelp-value of 0.0001 were obtained. Which suggest that the modelwas highly signicant and indicates only a 0.01% chance thatthe model F-value was larger due to noise.

While Table 5 shows that a model F-value of 6.14 and amodel p-value of 0.0129 were obtained, this predicts a 1.29%

Fig. 5 (A and B) 3D surface responses of Y1 (A) (peak area) and Y2. (B) (retention

This journal is ª The Royal Society of Chemistry 2013

chance that the model F-value was larger due to noise. Thisindicates that the predicted value of both responses were inapproximate to each other.

Fig. 4 shows the effect of individual factors on responses (Y1and Y2). The factors X1 and X2 show the effect on both responseswhile factor X3 has no effect. Further interaction of the factorswas checked for both responses which suggest that X1 (% AFbuffer) and X2 ow rate interact and affect the responses. Fig. 4Aand B

The one factor interaction study showed that the composi-tion of the mobile phase i.e. X1 (%AF buffer) was involved ininteraction while the two factor interaction study revealed that

time).

Analyst, 2013, 138, 1581–1588 | 1587

Analyst Paper

Publ

ishe

d on

17

Dec

embe

r 20

12. D

ownl

oade

d by

Uni

vers

ity o

f So

uthe

rn C

alif

orni

a on

02/

04/2

014

17:3

5:13

. View Article Online

both X1 and X2 gave a mixed response but X3 was not involved inany interaction. Hence pH was found to be the actual factorwhich can be easily evaluated from the contour plot. Fig. 5shows the 3D surface responses of Y1 and Y2.

4 Conclusions

The proposed method is simple, specic, accurate, linear andvalidated for the determination of lenalidomide in humanplasma over a range of 2–1000 ng mL�1. This offers a rapidand simple sample preparation which can facilitate the bio-study of lenalidomide. The method is stability indicating andit can be used for the routine as well as pharmacokineticanalysis of samples. The results obtained from the validationof method were satisfactory. The regulatory requirements foraccuracy, precision, sensitivity, selectivity, stability andruggedness were followed and achieved. The applied BBDdesign (32-Factorial) for optimisation of robustness parame-ters of validation that a minor changes in ow rate and mobilephase composition altered the responses i.e. peak area andretention time.

Conflict of interest

There are no conicts of interest.

Acknowledgements

The authors are very thankful to the management of FortisClinical Research Limited, Faridabad for supporting this work.

References

1 J. B. Bartlett, K. Dredge and A. G. Dalgleish, The evolution ofthalidomide and its IMiD derivatives as anticancer agents,Nat. Rev. Cancer, 2004, 4, 314–322.

2 P. G. Richardson, S. Jagannath, R. Schlossman,S. Zeldenrust, S. V. Rajkumar, M. Alsina, K. R. Desikan,C. S. Mitsiades, K. Kelly, D. Doss, M. McKenney, S. Gorelik,D. A. Warren, A. Freeman, R. Rich, R. Knight,M. Olesnyckyj, J. Patin, J. Zeldis, W. Dalton andK. C. Anderson, Blood, 2003, 102, 825.

3 K. C. Anderson, Lenalidomide and thalidomide:mechanisms of action-similarities and differences, Semin.Hematol., 2005, 42, S3–S8.

4 K. Dredge, R. Horsfall, S. P. Robinson, L. H. Zhang, L. Lu,Y. Tang, M. A. Shirley, G. Muller, P. Schafer, D. Stirling,A. G. Dalgliesh and J. B. Bartlet, Orally administeredlenalidomide (CC-5013) is anti-angiogenic in vivo andinhibits endothelial cell migration and Aktphosphorylation in vitro, Microvasc. Res., 2005, 69, 56–63.

1588 | Analyst, 2013, 138, 1581–1588

5 A. A. Giagounidis, U. Germing and C. Aul, Biological andprognostic signicance of chromosome 5q deletions inmyeloid malignancies, Clin. Cancer Res., 2006, 12, 5–10.

6 A. Naing, L. Sokol and A. F. List, Developmental therapeuticsfor myelodysplastic syndromes, J. Natl. Compr. CancerNetwork, 2006, 1, 78–82.

7 A. A. Giagounidis, U. Germing, C. Strupp, B. Hildebrandt,M. Heinsch and C. Aul, Prognosis of patients with del(5q)MDS and complex karyotype and the possible role oflenalidomide in this patient subgroup, Ann. Hematol.,2005, 84, 569–571.

8 A. List, S. Kurtin, D. J. Roe, A. Buresh, D. Mahadevan,D. Fuchs, L. Rimsza, R. Heaton, R. Knight and J. B. Zeldis,Efficacy of lenalidomide in myelodysplastic syndromes, N.Engl. J. Med., 2005, 352, 549–557.

9 K. McClellan and B. Jarvis, Desloratadine. Drugs., 2001, 61,789–796.

10 L. Liu, M. Qi, P. Wang and H. J. Li, High-performance liquidchromatographic method for the bioequivalence evaluationof desloratadine fumarate tablets in dogs, J. Pharm. Biomed.Anal., 2004, 34, 1013–1019.

11 F. C. Sutherland, A. D. De Jager, D. Badenhorst, T. Scanes,H. K. Hundt, K. J. Swart and A. F. Hundt, Sensitive liquidchromatography-tandem mass spectrometry method forthe determination of loratadine and its major activemetabolite descarboethoxyloratadine in human plasma, J.Chromatogr., A, 2001, 914, 37–43.

12 T. M. Tohniya, K. Hwang, E. R. Lepper, H. A. Fine,W. L. Dahut, J. Venitz, A. Sparreboom and W. D. Figg,Determination of CC-5013, an analogue of thalidomide, inhuman plasma by liquid chromatography–massspectrometry, J. Chromatogr., B: Anal. Technol. Biomed. LifeSci., 2004, 811, 135–141.

13 M. E. Bosch, A. J. R. Sanchez, F. S. Rojas and C. B. Ojeda,Recent advances in analytical determination ofthalidomide and its metabolites, J. Pharm. Biomed. Anal.,2008, 46, 9–17.

14 L. A. Svensson, J. Donnecke, K. E. Karlsson, A. Karlsson andJ. Vessman, Vancomycin-based chiral stationary phases formicro-column liquid chromatography, Chirality, 1999, 11,121–128.

15 S. K. Teo, R. S. Chandula, J. L. Harden, D. I. Stirling andS. D. Thomas, Sensitive and rapid method for thedeterminationof thalidomide in human plasma and semenusing solid-phase extraction and liquid chromatography–tandem mass spectrometry, J. Chromatogr., B: Anal.Technol. Biomed. Life Sci., 2002, 767, 145–151.

16 S. Beg, K. Kohli, S. Swain and M. S. Hasnain, Developmentand validation of RP-HPLC method for quantitation ofamoxicillin trihydrate in bulk and pharmaceuticalformulation using Box–Behnken experimental design, J.Liq. Chromatogr. Relat. Technol., 2012, 35, 393–406.

This journal is ª The Royal Society of Chemistry 2013