To quantify the small-molecule kinase inhibitors ceritinib, dacomitinib, lorlatinib, and nintedanib in human plasma by liquid chromatography/triple-quadrupole mass spectrometry

(1)

ContentslistsavailableatScienceDirect

Journal

of

Pharmaceutical

and

Biomedical

Analysis

jo u r n al ho me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / j p b a

To

quantify

the

small-molecule

kinase

inhibitors

ceritinib,

dacomitinib,

lorlatinib,

and

nintedanib

in

human

plasma

by

liquid

chromatography/triple-quadrupole

mass

spectrometry

G.D.

Marijn

Veerman

a,∗

_,

_Peter

_de

_Bruijn

a

_,

_Anne-Marie

_C.

_Dingemans

b

_,

Ron

H.J.

Mathijssen

a

_,

_Stijn

_L.W.

_Koolen

a,c

a_Department_of_Medical_Oncology,_Erasmus_University_Medical_Centre,_Rotterdam,_The_Netherlands b_Department_of_Pulmonology,_Erasmus_University_Medical_Centre,_Rotterdam,_The_Netherlands c_Department_of_Pharmacy,_Erasmus_University_Medical_Centre,_Rotterdam,_The_Netherlands

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received16September2020

Receivedinrevisedform20October2020 Accepted23October2020

Availableonline2November2020 Keywords:

Small-moleculekinaseinhibitor(SMKI) Ceritinib

Dacomitinib Lorlatinib Nintedanib

Ultra-performanceliquidchromatography tandemmassspectrometry(UPLC-MS/MS)

a

b

s

t

r

a

c

t

Multiplesmall-moleculekinaseinhibitorswithspeciﬁcmoleculartargetshaverecentlybeendeveloped forthetreatmentofcancer.Thisarticlereportsthedevelopmentandvalidationofanultra-performance liquidchromatography/tandemmassspectrometry(UPLC–MS/MS)methodtosimultaneouslyanalyse thesmall-moleculekinaseinhibitorsdacomitinib,ceritinib,lorlatinib,andnintedanibinhumanplasma. Forchromatographicanalyteseparation,anAcquityUPLC®_BEH_C18_column_1.7_␮m,₅₀_mm_x_2.1_mm

wasusedwithabinarygradientofpurewater/formicacid/ammoniumformate(100:0.1:0.02,v/v/v)and methanol/formicacid(100:0.1,v/v).Calibrationcurvesforallsmall-moleculekinaseinhibitorswere 5.00–500ng/mL.ValidationofthismethodmetallrequirementsoftheFoodandDrug administra-tion.Additionally,clinicalapplicabilitywasdemonstratedbyquantiﬁcationofmultiplesamplesfrom apharmacokineticstudyinpatientswithlungcancer.

1. Introduction

Small-moleculekinaseinhibitors(SMKIs)areoneofthe

back-bonesinthetreatmentofcancer.Theyaredevelopedtospeciﬁcally

blockdysregulated proteintyrosine kinasesintumour cells. [1]

TreatmentwithSMKIsisstandardofcareforvarioustypesofsolid

andhematologicmalignancies.Overthelastyears,theirnumber

hasvastlyincreased.Becauseoftheirspeciﬁcity,mostSMKIsare

prescribedbasedonmolecularcharacteristicsofatumour.

Innon-smallcelllungcancer(NSCLC),themostfrequent

molec-ulardriveraberrationsareepidermalgrowthfactorreceptor(EGFR)

mutationsandanaplasticlymphomakinase(ALK)translocations.

Together,theyrepresent10–20%ofallNSCLCpatients,withhigher

incidencesinAsiathanintherestoftheworld.[2]Dacomitinibisan

oralSMKIregisteredforEGFRmutatedNSCLC[3].Bothceritiniband

lorlatinibareSMKIsspeciﬁcallytargetingALKmutationsinNSCLC

∗ Correspondingauthorat:DepartmentofMedicalOncology,Laboratoryof Trans-lationalPharmacology,ErasmusUniversityMedicalCentre,PObox2040,3000CB, Rotterdam,TheNetherlands.

E-mailaddress:g.veerman@erasmusmc.nl(G.D.M.Veerman).

[3]. AllthreeSMKIsareadministeredoncedaily[3].Nintedanib

isavascularendothelialgrowthfactor(VEGF)receptorinhibitor

approvedincombinationwithdocetaxelforsecondlinetreatment

ofpatientswithNSCLC.Thistreatmenthowever,isindependent

ofmutationalstatusandisdosedinatwice-dailyscheduleof200

mg.[3]CombinationsofthesefourSMKIshavenotbeenreported

andbasedonthespeciﬁcindicationsofeachdrug,combination

treatmentisnotexpected.

Alsoinotherﬁeldsofmedicine,SMKIsareregisteredfor

progres-sivediseasesinwhichcellproliferationisunwanted.Nintedanibfor

example,hasanadditionalapprovalfortreatmentofprogressive

lungﬁbrosis.Contrarytoitsuseinmedicaloncology,nintedanibis

registeredasmonotherapyandadministered150mgtwicedaily.

[3]

MultipleSMKIsareknowntohavespeciﬁctargetplasma

con-centrationsandanarrowtherapeuticwindow.[4]Eventhoughoral

administrationofSMKIshasitsbeneﬁts,itadditionallyincreases

theintra-and interpatient variability. Due tointeractions with

coadministereddrugs[5],food,andbeverages[6,7]drug

absorp-tionandmetabolismcouldbealtered.Subsequentincreaseindrug

concentrationabovethetherapeuticwindowcouldleadto(severe)

https://doi.org/10.1016/j.jpba.2020.113733

(2)

toxicity,whereeffectivenessoftreatmentiscompromisedwhen

thetargetconcentrationisnotreached.

In order to investigate the therapeutic window of SMKIs,

compounds that could interact with them, and ultimately

performing therapeutic drug monitoring (TDM), a liquid

chro-matography/tandem mass spectrometric (LC–MS/MS) assay for

drugquantiﬁcationhastobedevelopedandvalidated.Thereare

multipleLC–MS/MSassayspublishedwhichseparatelyreportthe

quantiﬁcationofceritinib,[8–12]dacomitinib[13],lorlatinib[14],

or nintedanib[15–17]in plasma.Fordacomitiniband lorlatinib

however, these methods were only developed for non-human

plasma. We here describe thedevelopment and full validation

accordingtoU.S.Food&DrugAdministrationguidelines[18] of

anLC–MS/MSmethodforthesimultaneousquantiﬁcationofthe

SMKIsceritinib,dacomitinib,lorlatinib,andnintedanibinhuman

plasma.

2. Materialandmethods

2.1. Chemicals

Ceritinib(C28H36ClN5O3S)wasacquiredfromTorontoResearch

Chemicals (North York, Canada), dacomitinib (C24H25ClFN5O2)

at Sigma-Aldrich (Saint Louis, USA), lorlatinib (C21H19FN6O2)

at MedChemExpress(MonmouthJunction,USA) andnintedanib

(C31H33N5O4C28H33N7O2) at LC Laboratories (Woburn, USA).

The purity of all SMKIs was 98 % or higher. Labeled

inter-nal standard dasatinib-d8 was purchased at Toronto Research

Chemicals (Toronto,Canada). Blanc humanlithium heparinized

plasma was bought at BioIVT (Seralab), (West Sussex, United

Kingdom). Dimethylsulfoxide (DMSO) and ammonium formate

(1 mol/L) was obtained from Sigma-Aldrich. Acetonitrile, pure

waterandmethanoloriginatedfromBiosolve(Valkenswaard,The

Netherlands).FormicacidwaspurchasedfromJTBaker

(Deven-ter, TheNetherlands).Thesechemicalswereof analytical grade

or higher. Nitrogen originated from LindeGas (Schiedam, The

Netherlands).

2.2. Calibration

Ceritinib,dacomitinib,lorlatinib,andnintedanibweredissolved

inDMSOtocreatethreestocksolutionsof1.00mg/mL(inter-stock

deviation<5%). OnewasusedtoprepareQualityControl(QC)

samplesandonewasusedtoformcalibrationcurvestandards.

Sep-arately,dasatinib-d6wasdissolvedinDMSOtomakeasolutionof

1.0mg/mL.Acetonitrilewashereafterusedtodilute10,000-foldto

createaworkingsolutionof100ng/mLdasatinib-d6.Calibration

standardswiththefourSMKIconcentrationsof5.00,12.5,25.0,

50.0,125,250,450and500ng/mLwerepreparedinhumanlithium

heparinizedplasma.QCsamplesweremadetoyieldﬁnal

concen-trationsof5.00(lowerlimitofquantiﬁcation;LLQ),15.0(QC-low),

100(QC-middle),400(QC-high)and8,000(QC-diluted)ng/mL.The

QC-dilutedsamplesweresubsequentlydilutedpriortoprocessing

in humanplasmatothelevelof400ng/mL.Allstocksolutions,

internalstandardsolution,andpoolsofQCsampleswerestoredat

T<-70◦C.InternalstandardworkingsolutionwaskeptatT<6◦C

foramaximumofthreemonths.

2.3. Samplepre-treatment

Twenty-ﬁve␮Lplasmaofeverysamplewascombinedwith100

␮Linternalstandard solutionina 1.5mLvial.Solutionswere5

smixedand10mincentrifugedat18,000g.Hereafter,50␮Lof

thesupernatantwastransferredtoa96-deepwellplate,towhich

100␮Lwater/formicacid/ammoniumformate(100:0.1:0.02,v/v/v)

was added. After 5 min shaking, the plate was positioned in

thechilledautosampler.Aliquotsof5␮Lwereinjectedontothe

LC–MS/MScolumn.

2.4. LC–MS/MSsystem

The ultra-performance LC–MS/MS system originated from

WatersChromatographyBV(Etten-Leur,theNetherlands)was

con-sistedofaWatersAcquityUPLCI-ClassBinarySolventManager

connectingaWatersXEVOTQ-SMicroDetector.

2.4.1. Dataprocessing

MassLynxversion4.1SP4wasusedforthecollecting,processing

anddisplayingofdata.TargetLynxsoftwarecalculatedand

quanti-ﬁedallsamples.

2.4.2. Chromatographicsettings

Forchromatographicanalyteseparation,theAcquityUPLC®_BEH

C18column1.7␮m,50mmx2.1mm(WatersChromatography

BV)waswarmedtoT=40◦C.Columnﬂowratewas0.350mL/min

withabinarygradientofmobilephaseAwhichconsistedofpure

water/formicacid/ammonium formate (100:0.1:0.02,v/v/v)and

mobilephaseBwhichconsistedofmethanol/formicacid(100:0.1,

v/v).Afterinjection, separation wasyieldbylinearlyincreasing

phaseBfrom15%to60%from1.00minto5.00min.In30s,mobile

phaseBincreasedto90%andafter2.50min,itlinearlydecreased

to15%in30s.Duringthelast1.50minoftheoverallprocessing

timeof10.0min,mobilephaseBremainedat15%.Hereafter,the

injectionneedlewaswashedwithastrongorganicmixture

con-tainingacetonitrile,methanol,2-propanol,waterandformicacid

(25:25:25:25:0.1,v/v/v/v/v)topreventcarry-over.

2.4.3. Tandemmassspectrometry

Tandemmassspectrometry(MS/MS)wasperformedinthe

elec-trospraypositiveionizationmode.Capillaryionvoltagewassetat

3.00kVandthesourcetemperaturewassetat150◦C.

Desolva-tiontemperaturewassetat650◦Cwhile nitrogenwasusedas

desolvationgasataﬂowrateof1,200l/h.CollisioncellPirani

pres-surewaskeptat∼4.5e-3mbarbyargongas.Masstransitionsof

ceritinib,dacomitinib,lorlatinibandnintedanibandtheinternal

standarddasatinib-d8wereperformedbydirectinfusionafter

dis-solvingallcompoundsinacetonitrile/water/formicacid(40:60:0.1,

v/v/v).Thesixmasspairswerequantiﬁedduringmultiplereaction

monitoring.TheMS/MSfunctionsweremanuallyadjustedtoobtain

theoptimalparameterswhicharepresentedinTable1.

2.4.4. Quantiﬁcation

PeakarearatiosperSMKIcompared totheinternalstandard

wereusedinordertofromcalibrationcurves;ceritinib(558>433),

dacomitinib(470>124),lorlatinib(407>228)andnintedanib(540

>113)todasatinib-d8(496>406)witha1/concentration2_weight

factor.

2.5. Validation

ThemethodhasbeenvalidatedaccordingtothelatestU.S.Food

&DrugAdministration(FDA)guidelines.[18]

2.5.1. Speciﬁcityandselectivity

Thespeciﬁcityandselectivityofthisassayweretestedatthe

levelof theLLQ(5.00 ng/mL)for which10 sets ofsingleblank

humanplasmawerespiked.Allsampleswereanalysed

simulta-neously.

2.5.2. Accuracy,precisionandmatrixeffect

FivesetsofLLQandQCsampleswereanalysedinthree

(3)

Table1

Massspectrometrysettingsformultiplereactionmonitoring.

Compound Parent(m/z) Daughter(m/z) Dwell(msec) Cone(V) Collision(V)

Ceritinib 558 433 54 23 29 Ceritinib 558 516 54 25 24 Dacomitinib 470 124 54 20 34 Lorlatinib 407 228 54 35 21 Nintedanib 540 113 54 10 28 Dasatinib-d8 496 406 54 30 27

Transitionsusedforthequantitationarepresentedinbold.

curvestandards.Tocalculatetheassay’saccuracy,within-runand

between-runprecision,aunidirectionalanalysisofvariance

(one-wayANOVAanalysis)wasrun,withthecycleasvariable.[19]The

matrixeffectwastestedintriplicatewiththepost-extraction

addi-tionmethod:sixblankplasmasamplesunderwentpreparationand

extraction,whereaftertheywerespikedatQClow(15.0ng/mL)and

QChigh(400ng/mL)forallSMKIs(A).Aneatsolventstandard

solu-tionservedasreference(B).Afteranalysingallsamples,thematrix

effectcouldbequantiﬁedbyB

Ax100%.

2.5.3. Stability

Stabilityofceritinib,dacomitinib,lorlatinibandnintedanibin

plasma was tested at concentration levels of QC-low, QC-high

and QC-diluted.QCsamples(intriplicate)werestored at

ambi-enttemperatureforathree-dayperiodaswellasfollowingthree

freeze-thawcyclesatwhichthesampleswerethawedina

water-bathat37◦Cfor30minandrefreezeforatleast18hat<-70◦C.

Autosamplerstabilitywastestedintriplicatebyrepeatedinjection

ofprocessedQC-samplesstoredinachilledautosamplerat10◦C.

2.6. Clinicalapplicability

Theassay’sclinicalapplicabilitywasdemonstratedbyanalysing

plasmasamplesfrompatientstreatedwithceritinib,lorlatiniband

nintedanib.Thesesampleswereprospectivelycollectedaspartofa

largepharmacokineticstudyinwhichpatientstreatedwithall

dif-ferentSMKIswereincluded.InordertomeasureallSMKIs,another

LC–MS/MSmethodhasalsobeendeveloped.[20]Thestudyhas

beenapprovedbythelocalethicscommitteeoftheErasmusMC

(MEC 16–643)and patientsprovided writteninformed consent

priortostudyenrolment.Samplesweretakenaftersixortwelve

weeksoftreatmenttoguaranteesteadystateplasma

concentra-tions.PatientswasaskedtowaitwithSMKIintakeuntilafterblood

withdrawal,inordertodetermineplasmatroughconcentrations.

Subsequentaftercollection,sampleswerecentrifugedfor10min

at2500gand4◦C,afterwhichtheplasmawaskeptat–70◦C.

Unfortunately, nopatientsweretreatedyetwithdacomitinibin

thisstudy.

2.7. Incurredsamplereanalysis

In linewith the clinicalapplicability of this assay, incurred

samplereanalysiswastestedtoensuretheprecisenessofstudy

samplequantiﬁcationinasecondrun.Therefore,allusedsamples

inparagraph2.6underwentindependentreanalysis.Hereafter,the

relativedifferencebetweentheﬁrstsampleconcentration(C1)and

thesecond-repeated-sampleconcentration(C2)wascalculated

by: (C1(C1–+C2)C2)

2

∗100%.

3. Resultsanddiscussion

3.1. Methoddevelopment

Combining the optimal mass-spectrometric parameters and

retention times with mobile phase A (pure water/formic

acid/ammoniumformate;100:0.1;0.02,v/v/v)and mobilephase

B(methanol/formicacid;100:0.1,v/v)atagradient0.350mL/min

ﬂowrate,purestpeaksforeverySMKIweredetected.Selected

tran-sitions,proposedfragmentationsandionspectrafortheseSMKIs

areshowninFig.1.Primarytosecondaryionratiosdemonstrated

thequalityforeveryobservedpeak.Theselectedproductionsare

presentedinTable1.Dasatinib-d8wasanacceptableinternal

stan-dardforallcompounds,whereotherauthorsreportusingother

(isotope-labelled)internalstandards.[8–17]Separationofanalyts

wasyieldedwithasteppedgradient(paragraph2.4.2).Herewith,

representativechromatogramsforallSMKIsandinternalstandard

werecreated,asillustratedinFig.2.Retentiontimesforceritinib,

dacomitinib,lorlatiniband nintedanibwere5.72, 3.61,3.33and

5.11minrespectively.Acetonitriledeproteinisationprovedtobea

successfulmethodforquantifyingceritinib,dacomitinib,lorlatinib

andnintedanibinhumanplasmaintherangeof5.00–500ng/mL.

Toavoidcarry-over,awashperiodof3minof90%eluensBwas

built-in,whichledtoatotalrun-timeof10min.

3.2. Assayperformance

3.2.1. Speciﬁcityandselectivity

Calibrationcurveregressioncoefﬁcients(R2₎_were_linear_from

5.00–500ng/mLandrangedfrom0.9967to0.9990.Thelowerlimit

of quantitation (5.00ng/mL) was validated in tenindependent

poolsofhumanplasmaandbyﬁvemeasurementsrepeatedin

trip-licate(15samplesintotal).ForeverySMKI,allanalysedLLQsamples

were in the acceptance ranges of accuracy of 80–120%. Blank

humanplasmadidneitherinterferewiththeSMKIsnortheinternal

standard.Meanceritinibconcentrationwas4.85ng/mL(standard

deviation;SD0.259ng/mL),meandacomitinibconcentrationwas

4.61ng/mL(SD0.303ng/mL),meanlorlatinibconcentrationwas

4.59ng/mL(SD0.273ng/mL),andmeannintedanibconcentration

was4.65ng/mL(SD0.214ng/mL).

3.2.2. Accuracy,precisionandmatrixeffect

Theassay’saccuracy,within-runandbetween-runprecisionat

concentrationlevelsofLLQandQClow,middle,highanddiluted

mettherequirementsofvalidationoftheFDA(LLQrange80–120

%,QCrange85–115%).[18]All15analysedsamplesforeverySMKI

ateveryconcentrationfellintheseacceptablevalidationranges.

Thepost-extractionadditionmethodwithsixindependent

sam-plesatQClowandhighshowednoevidenceforsigniﬁcantmatrix

effectfor anySMKI.Thematrixeffectrangedfrom84to100%,

showingnointerferenceofthematrixinthismethod.Theresultsof

accuracy,within-runandbetween-runprecisionandmatrixeffect

(4)

Fig.1. Selectedtransitions,proposedfragmentationsandionspectraforceritinib(A),dacomitinib(B),lorlatinib(C)andnintedanib(D).

Table2

Accuracy,within-runprecision,between-runprecisionandmatrixeffect.

Compound Spiked(ng/mL) GrandMean(ng/mL) AverageAccuracy(%) WRP(%) BRP(%) Matrixeffect1_(%)

Ceritinib LLQ 5.00 5.44 108.8 12.7 # – QCLOW 15.0 15.5 103.3 3.13 5.82 81.9±2.38 QCMIDDLE 100 104 104.0 2.62 5.36 – QCHIGH 400 408 102.0 2.44 4.29 84.1±3.56 QCDILUTED 8000 7787 97.3 3.01 3.20 – Dacomitinib LLQ 5.00 5.24 104.8 2.49 2.35 – QCLOW 15.0 14.7 98.0 2.92 1.28 94.7±2.20 QCMIDDLE 100 95.9 95.9 1.17 0.67 – QCHIGH 400 379 94.8 2.18 1.67 97.9±3.13 QCDILUTED 8000 7308 91.4 3.52 0.94 – Lorlatinib LLQ 5.00 4.80 96.0 2.22 1.20 – QCLOW 15.0 15.4 102.7 2.16 # 103±1.19 QCMIDDLE 100 103 103.0 1.65 0.47 – QCHIGH 400 381 95.3 1.80 2.04 100±0.914 QCDILUTED 8000 7509 93.9 3.39 # – Nintedanib LLQ 5.00 4.89 97.8 2.14 2.30 – QCLOW 15.0 14.4 96.0 2.02 2.53 99.9±1.08 QCMIDDLE 100 95.7 95.7 1.62 1.06 – QCHIGH 400 373 93.3 2.44 2.30 98.4±2.40 QCDILUTED 8000 7199 90.0 3.14 2.54 –

Abbreviations:WRP=within-runprecision;BRP=between-runprecision.#=noadditionalvariationobservedbyperformingtheassayindifferentruns.1₌_data_represent meanvalues±standarddeviation(n=3).

3.2.3. Stability

AllSMKIswerestablewhenthetestedhumanplasmasamples

wherekeptforthreedaysatambienttemperature,keptfor23hin

theautosamplerandfreezedandthawedthreesubsequenttimes.

Everysamplefellwithin15%ofnominalconcentrations(whichis

(5)

Fig.2. Representativechromatogramsofdoubleblankprocessedplasmasamples(A,D,GandJ),plasmasamplesspikedattheconcentrationoftheLLQforceritinib(B), dacomitinib(E),lorlatinib(H),andnintedanib(K),andplasmasamplescollectedinthepharmacokineticstudycontaining875ng/mLceritinib(C),305ng/mLlorlatinib(I), 37.3ng/mLnintedanib(L)andinternalstandarddasatinib-d8(F).

Long-termstabilitywastestedatthetimethisreportedwas

writ-ten.SincetheceritinibstockwasstoredearlierthantheotherSMKI

stocks,itslong-termstabilitywasatleast2yearsand8months.For

dacomitinib,lorlatinibandnintedanib,long-termstabilitywasat

(6)

Table3

Stabilityafterdifferentstorageconditions.

Reference 3daysAmbientTemp1 ₃_{freeze-thaw-cycles}1 _Autosampler₍₂₃_h)2

Ceritinib QCLOW 16.1±0.529 14.3±0.153(89.0%) 16.1±0.458(100%) 16.2(1.3%) QCHIGH 434±8.08 403±16.6(92.8%) 429±14.7(98.8%) 417(96.3%) QCDILUTED 8,214±180 8,436±160(1.3%) 9443±279(115%) 8575(98.9%) Dacomitinib QCLOW 15.0±0.100 12.1±0.058(80.9%) 13.6±0.100(90.7%) 14.1(97.9%) QCHIGH 392±8.96 343±3.22(87.3%) 387±7.94(98.6%) 374(95.7%) QCDILUTED 7,429±376 7,112±267(95.7%) 7834±898(106%) 7482(95.8%) Lorlatinib QCLOW 15.9±0.611 15.6±0.173(98.3%) 15.7±0.100(98.9%) 15.3(96.2%) QCHIGH 396±11.0 396±4.04(100%) 396±6.43(99.8%) 375(91.2%) QCDILUTED 7,770±405 8,300±266(107%) 8521±107(110%) 7725(93.3%) Nintedanib QCLOW 14.8±0.379 14.1±0.208(94.8%) 14.4±0.493(96.9%) 15.1(97.4%) QCHIGH 382±12.8 378±13.9(98.9%) 387±15.6(104%) 392(94.9%) QCDILUTED 7,477±112 8,140±175(109%) 8462±286(1.3%) 8020(97.4%)

Datarepresentmean±standarddeviation(n=3).1₌_versus_value_the_reference.2₌_versus_value_“0h”_(n₌_1).

Table4

Incurredsamplesreanalysis.

Compound Firstanalysis Secondanalysis Mean %Difference Conc.(ng/mL) Conc.(ng/mL) Conc.(ng/mL) (%) Ceritinib 1 818 875 847 −6.7 2 922 911 917 1.2 3 195 192 194 1.6 4 670 638 654 4.9 5 855 838 847 2.0 Nintedanib 1 15.0 15.2 15.1 −1.3 2 35.8 37.3 36.6 −4.1 3 11.2 12.1 11.7 −7.7 4 7.41 7.83 7.62 −5.5 Lorlatinib 1 300 305 303 −1.7 2 631 632 632 −0.2 3 335 361 348 −7.5 4 188 191 190 −1.6 5 36.0 38.7 37.4 −7.2 6 287 395 291 −2.7

%Difference=(ﬁrstanalysis–secondanalysis)/(meanofﬁrstandsecond analy-sis)*100.

3.3. Clinicalapplicability

Fifteenstudysamplesweresuccessfullyusedtoprovethe

clini-calapplicabilityofthisassay.Respectivelyﬁve,sixandfourplasma

samplesofpatientstreatedwithceritinib,lorlatinibandnintedanib

wereprocessedandquantiﬁed.Table4presentsthemeasured

con-centrations,whichwereallwithinthisassay’sconcentrationrange.

Onlyceritinibandnintedanibhaveexploratoryresults

indicat-ingtowardspotentialtargetconcentrationswhichcouldbeused

fortherapeuticdrugmonitoring(TDM).[4]WithTDM,dose

opti-mization and personalization couldlead to higher efﬁcacyand

lowertoxicityoftreatmentwithSMKIs.Furtherpharmacokinetic

researchshouldthereforebefocusingonwhetherornotthereexist

solidtargetconcentrationsfortreatmentefﬁcacyortoxicity.This

isnecessaryinordertouseplasmadrugconcentrationsfor

opti-mizationoftreatmentwiththeseSMKIsinthefuture.

3.4. Incurredsamplereanalysis

Reanalysingthepatient samplesdidnotresult insigniﬁcant

deviation inquantiﬁcation (lastcolumns ofTable 4.). Henceall

SMKIspassedtheincurredsamplereanalysistest,inwhichtheFDA

criteriaforsmall-moleculesisthatatleasttwo-thirdsofthe

sam-plesshouldhavenomorethan20%differencebetweeneachother.

[18]Ceritinibhadanaverage3.3%difference(-6.7to4.9%),which

wasforlorlatinib3.5%(-7.5to-0.2%)andfornintedanib4.7%(-7.7

to-1.3%).

3.5. Performancecomparisonwithothermethods

TherearemultipleLC MS/MSassayspublishedwhichreport

thequantiﬁcationofceritinib,[8–12]dacomitinib[13],lorlatinib

[14],ornintedanib[15–17]inplasma,withsimilarquantiﬁcation

concentrationranges.Forbothceritinibandnintedanib,

method-ologiesaredescribedtoquantifymultipleSMKIssimultaneously

inhumanplasma[10,16].However,onlysingleSMKILC MS/MS

assayshavebeendescribedforquantifyingdacomitinib(LLQ1.00

ng/mL)and lorlatinib (LLQ2.00 ng/mL)in non-human (rodent)

plasma.[13,14]Incomparisontotheselattermethodologies,the

herereportedmethodismoreefﬁcienttoquantifyessentialSMKIs

simultaneouslyanditistheﬁrstthatisvalidatedforhumanplasma.

It isimportant toextrapolateanimal studiestohumanstudies,

since differences in validation and applicability couldoccur. In

themulti-assaysforceritinibandnintedanib, multiple

(isotope-labelled)internalstandardsareused.Acompoundhaspotentialto

serveasaninternalstandardwhenishas(almost)equalchemical

characteristics(e.g.boilingtemperatureandatomicgroups)

com-paredtothecompoundsthathavetobequantiﬁed.Itisimportant

thatcompoundwillnotbepresentintheplasmasamplesthough.

Therefore,anisotope-labelledinternalstandardisoftenusedfor

everyanalyte.However,usingoneinternalstandardoverusingall

differentinternalstandardsis besidesﬁnanciallyattractive,also

moreefﬁcientinpreparationandstorage,andreducesthe

inﬂu-enceofrepeatedrandomerrorsinpreparationandquantiﬁcation

oftheinternalstandards.Inthismethod,dasatinib-d8was

success-fullyusedasinternalstandardbecauseofitsverysimilarchemical

characteristics.

4. Conclusion

This article reports the development and validation of the

ﬁrstultra-performanceliquidchromatography/tandemmass

spec-trometry (UPLC–MS/MS)methodtosimultaneously analysethe

SMKIsdacomitinib,ceritinib,lorlatinib,andnintedanibinhuman

plasma,withdasatinib-d8 assoleinternal standard. Calibration

curvesforallSMKIswerelinearintherangeof5.00–500ng/mL.

withalowerlimitofquantitationof5.00ng/mLforallcompounds.

ValidationofthismethodmetallrequirementsoftheFDA.

Addi-tionally,clinicalapplicabilitywasdemonstratedbyquantiﬁcation

(7)

patients.ThismethodcanhencebeusedinthefutureforTDMina

clinicalsetting.

Financialsupport

Pﬁzercontributedﬁnanciallytothisworkwithanunrestricted

grant(contractnumber54332551).

CRediTauthorshipcontributionstatement

G.D. Marijn Veerman: Conceptualization, Methodology,

Resources,Writing-originaldraft,Visualization,Funding

acquisi-tion.PeterdeBruijn:Conceptualization,Methodology,Validation,

Formal analysis, Investigation, Resources, Writing - review &

editing, Visualization, Project administration. Anne-Marie C.

Dingemans: Resources, Writing - review & editing, Funding

acquisition.RonH.J.Mathijssen:Resources,Writing -review&

editing,Fundingacquisition.StijnL.W.Koolen:Resources,Writing

- review&editing,Supervision, Projectadministration,Funding

acquisition.

DeclarationofCompetingInterest

None.

References

[1]D.S.Krause,R.A.VanEtten,Tyrosinekinasesastargetsforcancertherapy,N. Engl.J.Med.353(2005)172–187.

[2]A.S.Tsao,G.V.Scagliotti,P.A.BunnJr.,D.P.Carbone,G.W.Warren,C.Bai,H.J. deKoning,A.U.Yousaf-Khan,A.McWilliams,M.S.Tsao,P.S.Adusumilli,R. Rami-Porta,H.Asamura,P.E.VanSchil,G.E.Darling,S.S.Ramalingam,D.R. Gomez,K.E.Rosenzweig,S.Zimmermann,S.Peters,S.H.IgnatiusOu,T. Reungwetwattana,P.A.Janne,T.S.Mok,H.A.Wakelee,R.Pirker,J.Mazieres, J.R.Brahmer,Y.Zhou,R.S.Herbst,V.A.Papadimitrakopoulou,M.W.Redman, M.W.Wynes,D.R.Gandara,R.J.Kelly,F.R.Hirsch,H.I.Pass,Scientiﬁcadvances inlungCancer2015,J.Thorac.Oncol.11(2016)613–638.

[3]FoodandDrugAdministration,.Drugs@FDA:FDA-approvedDrugs,2020 https://www.accessdata.fda.gov/scripts/cder/daf/.

[4]R.B.Verheijen,H.Yu,J.H.M.Schellens,J.H.Beijnen,N.Steeghs,A.D.R.Huitema, Practicalrecommendationsfortherapeuticdrugmonitoringofkinase inhibitorsinoncology,Clin.Pharmacol.Ther.102(2017)765–776.

[5]K.Hussaarts,G.D.M.Veerman,F.G.A.Jansman,T.vanGelder,R.H.J.Mathijssen, R.W.F.vanLeeuwen,Clinicallyrelevantdruginteractionswithmultikinase inhibitors:areview,Ther.Adv.Med.Oncol.11(2019),1758835918818347. [6]G.D.M.Veerman,K.Hussaarts,F.G.A.Jansman,S.W.L.Koolen,R.W.F.van

Leeuwen,R.H.J.Mathijssen,Clinicalimplicationsoffood-druginteractions withsmall-moleculekinaseinhibitors,LancetOncol.21(2020)e265–e279. [7]G.D.M.Veerman,K.Hussaarts,R.Peric,E.Oomen-deHoop,K.D.Landa,C.H. vanderLeest,S.D.Broerse,H.B.Rutten,H.N.A.Belderbos,C.M.J.Steendam,

M.S.Paats,S.L.W.Koolen,A.C.Dingemans,T.vanGelder,R.W.F.vanLeeuwen, J.Aerts,R.H.J.Mathijssen,Inﬂuenceofcow’smilkandesomeprazoleonthe absorptionoferlotinib:arandomized,crossoverpharmacokineticstudyin lungCancerpatients,Clin.Pharmacokinet.(2020).

[8]O.Heudi,D.Vogel,Y.Y.Lau,F.Picard,O.Kretz,Liquidchromatographytandem massspectrometrymethodforthequantitativeanalysisofceritinibinhuman plasmaanditsapplicationtopharmacokineticstudies,Anal.Bioanal.Chem. 406(2014)7389–7396.

[9]C.Lanshoeft,O.Heudi,M.Raccuglia,L.A.Leuthold,F.Picard,O.Kretz,Ultrafast quantitativeMS-basedmethodforceritinibanalysisinhumanplasma samplesfromclinicaltrial,Bioanalysis7(2015)425–435.

[10]M.Herbrink,N.deVries,H.Rosing,A.D.R.Huitema,B.Nuijen,J.H.M.Schellens, J.H.Beijnen,Developmentandvalidationofaliquidchromatography-tandem massspectrometryanalyticalmethodforthetherapeuticdrugmonitoringof eightnovelanticancerdrugs,Biomed.Chromatogr.32(2018).

[11]X.Bao,J.Wu,N.Sanai,J.Li,Aliquidchromatographywithtandemmass spectrometrymethodforquantitatingtotalandunboundceritinibinpatient plasmaandbraintumor,J.Pharm.Anal.8(2018)20–26.

[12]P.Du,Y.Guan,Z.An,P.Li,L.Liu,AselectiveandrobustUPLC-MS/MSmethod forthesimultaneousquantitativedeterminationofanlotinib,ceritiniband ibrutinibinratplasmaanditsapplicationtoapharmacokineticstudy,Analyst 144(2019)5462–5471.

[13]X.Qiu,Q.Lin,Z.Ning,X.Qian,P.Li,L.Ye,S.Xie,Quantitativebioanalytical assayforthehumanepidermalgrowthfactorreceptor(HER)inhibitor dacomitinibinratplasmabyUPLC-MS/MS,J.Pharm.Biomed.Anal.166 (2019)66–70.

[14]C.Spatari,W.Li,A.H.Schinkel,G.Ragno,J.H.M.Schellens,J.H.Beijnen,R.W. Sparidans,BioanalyticalassayforthequantiﬁcationoftheALKinhibitor lorlatinibinmouseplasmausingliquidchromatography-tandemmass spectrometry,J.Chromatogr.BAnalyt.Technol.Biomed.LifeSci.1083(2018) 204–208.

[15]D.Lin,L.M.Qiao,Y.N.Zhang,Y.Liu,X.S.Liu,Simultaneousdeterminationof nintedanibanditsmetabolitebyUPLC-MS/MSinratplasmaandits applicationtoapharmacokineticstudy,J.Pharm.Biomed.Anal.117(2016) 173–177.

[16]R.Reis,L.Labat,M.Allard,P.Boudou-Rouquette,J.Chapron,A.Bellesoeur,A. Thomas-Schoemann,J.Arrondeau,F.Giraud,J.Alexandre,M.Vidal,F. Goldwasser,B.Blanchet,Liquidchromatography-tandemmassspectrometric assayfortherapeuticdrugmonitoringoftheEGFRinhibitorsafatinib, erlotinibandosimertinib,theALKinhibitorcrizotinibandtheVEGFRinhibitor nintedanibinhumanplasmafromnon-smallcelllungcancerpatients,J. Pharm.Biomed.Anal.158(2018)174–183.

[17]J.M.Janssen,N.deVries,N.Venekamp,H.Rosing,A.D.R.Huitema,J.H.Beijnen, Developmentandvalidationofaliquidchromatography-tandemmass spectrometryassayfornineoralanticancerdrugsinhumanplasma,J.Pharm. Biomed.Anal.174(2019)561–566.

[18]F.D.A.U.S,GuidanceforIndustry,BioanalyticalMethodValidation,2018 www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/ Guidances/UCM070107.pdf.

[19]H.Rosing,W.Y.Man,E.Doyle,A.Bult,J.H.Beijnen,Bioanalyticalliquid chromatographicmethodvalidation.Areviewofcurrentpracticesand procedures,JLiqChromatogrRT23(2000)329–354.

[20]G.D.M.Veerman,M.H.Lam,R.H.J.Mathijssen,S.L.W.Koolen,P.deBruijn, Quantiﬁcationofafatinib,alectinib,crizotinibandosimertinibinhuman plasmabyliquidchromatography/triple-quadrupolemassspectrometry; focusingonthestabilityofosimertinib,J.Chromatogr.BAnalyt.Technol. Biomed.LifeSci.1113(2019)37–44.