D3S-0o1in advanced solid tumors with KRAsG2c mutations:a phase 1trial

Received:6March2025

Accepted:2April2025

Published online:29April 2025

Byoung Chul Cho 115,16, Shun Lu215, Myung Ah Lee?, Zhengbo Song, JohnJ.Park,SunMinLim',ZimingLi²,JunZhao6,GaryRichardson, Yanqiao Zhang°,Jun Zhang?,Anwen Liu',Herbert H. Loong12,Cheng Chen3, Jia Wang,YandongShen,ZifeiFan3,QianChen,HuiWang,JingZhang3, Zhi Jian Chen13.15, Melissa L. Johnson14 & Tony Mok@ 1.12,16

D3S-001is a next-generation KRAS-G12Cinhibitor(G12Ci) designed to enhance targetengagementefficiency and overcome growthfactor-induced nucleotideexchange.D3s-001wasevaluatedinaphaseladose-escalation study in patients with advanced solid tumors harboring K R A S^{G12C} mutation (N=42) and a phase 1b expansion cohort of patients with non-small-cell lung cancer (NsCLC) whose disease progressed after prior G12Citherapy (N=20) The primaryendpoints weresafety and determination ofthemaximum tolerated dose.Secondaryendpoints included pharmacokinetics, confirmed objective response rate(ORR)and disease control rate.D3s-001 demonstrated dose-dependent pharmacokinetics and no dose-limiting toxicities,andthemaximumtolerated dosewasnotreached.Grade3 treatment-related adverseeventswerereportedinsevenpatients (16.7%) intheG12Ci-naivedose-escalationcohortandtwopatients (10.0%) inthe G12Ci-pretreatedNsCLCexpansioncohort.Therewere nograde4or5 treatment-related adverse events.D3s-001600mgwas selected asthe dose forfurtherinvestigationbasedonpharmacokinetics.Confirmed ORRin the G12Ci-naive population was 73.5% overall (25 of34),and 66.7% (14 of21), 88.9% (8of9)and 75.0% (3of4)inpatientswithNSCLC,colorectalcancer andpancreaticductaladenocarcinoma,respectively.Amongpatientswith G12Ci-pretreatedNSCLC,ORRwas 30.0% (6of20)anddiseasecontrolrate was 80.0% (16of20).Thisstudydemonstratesthesafetyandtolerability of D3S-001 monotherapy with promising antitumor activity. The phase 1b expansionphaseis ongoing.ClinicalTrials.govidentifier:NCT05410145.

Kirstenratsarcomaviraloncogenehomolog(KRAS)genemutations are commononcogenicdriversinsolid tumorsl2.Approximately 80% ofKRAS mutationsinvolve alterationsin codon12(refs.1,2), andKRASglycine-to-cysteine(G12C)mutationisfoundin11- 15% of patients with NSCLC, 3-4% ofpatients withcolorectal cancer(CRC) and 1-2% of patients with pancreaticcancer3-6.KRASfunctions asa molecularswitch,cyclingbetweentheinactiveguanosinediphosphate (GDP)-bound and theactive guanosinetriphosphate(GTP)-bound state, regulatingdownstreamsignalingpathwaysthatcontrolcellsurvival and proliferation78.Codon 12 mutations impair GTPase-activating protein-facilitated hydrolysis,leadingtoKRASoveractivation and drivingcancerinitiationandprogression.Unlikeothercodon12mutations,G12C-mutatedKRASretainsthemajorityofintrinsichydrolysis activity;therefore,targetingtheGDP-boundformisaviabletherapeutic strategy91o.Thisapproachisparticularlyeffective whenusingcovalent chemistrytoirreversiblyconjugateasmall-moleculeinhibitorwith themutantcysteine,therebylockingKRASintheGDP-boundinactive conformation10-2.First-generationKRAS-G12Cinhibitors(G12Ci),such asadagrasibandsotorasib,demonstratedclinicalproofofconceptfor thisstrategyandreceived accelerated approvalbytheUSFood and DrugAdministrationasmonotherapiesforsecond-linetreatmentof K R A S^{G12C} mutant NSCLC.In addition, adagrasib combined with the anti-epidermalgrowthfactorreceptor(EGFR)antibodycetuximab, andsotorasibcombinedwithpanitumumab,areapprovedforthe treatmentof K R A S^{G12C} mutant CRC.However,the clinical benefits of first-generationKRAS-G12Ciremainmodestinthemagnitudeand durationofresponses.InNSCLC,ORRswithadagrasib andsotorasib monotherapyare 31.9% and 28.1% withmedianresponsedurations of8.3and8.6months,respectively13.14.InCRC,benefitsaremore limited,with ORRsrangingfrom 9.7% to 23% andamedianresponse durationof4.3 months15.16.Consequently,it is necessary to combine KRAS-G12CiwithEGFRantibodiesto address epidermalgrowthfactor(EGF)-mediated reactivation ofKRAS and mitogen-activated proteinkinase(MAPK) pathway signalingin CRC.ORRs of 34% werereportedforadagrasibcombinedwithcetuximaband 26% for sotorasibcombined withpanitumumab16-1s.Moreeffectivetreatment optionsforpatientswithtumorsharboring KRAsGi2cmutations remain an unmetneed.

To guidethe development of next-generationKRASinhibitors andinformrationalcombinationstrategies,itiscriticaltounderstandwhetherthemodestclinicalbenefitsoffirst-generationG12Ci relatetoKRAStargetordiseasebiology,potencyorexposure ofthe compounds,orboth.Asidefrom thebiological heterogeneity of KRASmutanttumors,questionsremainregardingwhetherfirstgenerationGi2Ciachieveoptimaltargetinhibitionatclinicallyrelevantdoses.Inpreviousanalysesoftargetengagementefficiency,free drugconcentrationsrequiredforsotorasib and adagrasibtoachieve completetargetengagementexceededtheirclinicalpharmacokinetic exposures19.Thisinsufficient targetinhibition may result inunoccupiedKRAS-G12C proteinsin the tumor,enabling continued tumor progressionandcontributingtolimitedresponse andrapiddisease progression.Inaddition,first-generationinhibitorsarehighlysusceptibletoEGF-stimulatednucleotideexchange,whichshiftsKRAS fromitsGDP-bound inactiveformtoits GTP-bound activeform20.21. Becausetheconformationofthedirectdrug-bindingpocketdiffers substantiallybetweentheGDP-andGTP-boundforms,EGF-stimulated nucleotideexchangefurtherdiminishestargetengagementefficiency of these compounds20.21.Thisis particularly relevant in CRC,where higherlevelsofreceptortyrosinekinase(RTK)activationareobserved comparedwithNSCLC22.Accordingly,the antitumoractivityofsotorasibandadagrasibissubstantiallyimprovedwhencombinedwith anti-EGFRantibodies,whichreducesEGFRactivation-mediatedKRAS GDP-to-GTPtransition16-18.Theseinsights underscore two critical challenges for next-generation KRAS-G12C inhibitors:(1) achieving completetargetcoverageatclinicallyrelevant dosestoensurethatfree drugexposureiswellabovetheconcentrationrequiredforcomplete targetengagement;and (2) overcominggrowthfactor-stimulated nucleotide exchangetoeffectively'trap'KRAS-G12Cproteinsintheir inactiveform,even inthe presence ofgrowth factors.

D3S-001is a next-generationGDP-boundKRAS-G12Cinhibitor designed toenhancetargetengagement efficiency and overcome growthfactor-inducednucleotideexchangel9.Inpreclinicalprofiling, D3S-o01demonstratedimprovedcovalentpotencyoveradagrasib andsotorasib,achievingcompletetargetinhibitionatsingle-digit nanomolarconcentrations-twoordersofmagnitudemorepotentthan sotorasibandadagrasib'.Moreover,thetargetengagementkineticsof D3S-001differsubstantiallyfromfirst-generationKRAS-G12Ci,demonstratinganexceptionallyfastrateinlockingKRASintoitsinactivestate, withthereactionhalf-life (t_{1/_{2}}) exceedingthereportedrateofGDPdissociationfromKRAS-G12C9.Consequently,D3S-001effectively depletes cellularactiveKRASandpreventsdownstreamMAPKactivation,even inthe presenceofgrowthfactors"9.D3s-001demonstrated consistent robustantitumoractivityinmultiplepreclinicalxenograftmodels, includingNSCLC,CRCandpancreaticductaladenocarcinoma(PDAC) models9Furthermore,strongantitumoreffectswereobservedinacell line-derivedxenograftmodelwithsotorasibresistanceduetoKRAsG12c gene amplification and in a patient-derived xenograft model from sotorasib-resistanttumors9.D3s-001wasfoundtopenetratethecentralnervoussystem andinducedurableintracranialtumorregression inmouse modelsofbrainmetastasis9.Giventhesepromisingfeatures, D3S-oo1ispredictedtoimproveclinicaloutcomesforpatientswith K R A S^{{G12C}} mutations,particularly intumor types or disease conditions lessresponsivetofirst-generationKRAS-G12Ci,includingpatients withCRC,patientswhohavebecomeresistanttofirst-generation inhibitorsbecause of insufficient target inhibition and patients who havedevelopedbrainmetastasis.

The first-in-human phase 1 and 2 trial (clinicaltrials.gov NCT05410145)of D3S-001was designed toinvestigate thesehypothesesinrelevantpatientpopulations.Wereportresultsfromthephase ladose-escalationpart,evaluatingD3s-o01inpatientswithadvanced solid tumors harboring a K R A S^{G12C} mutation.The primary objective wastoassesssafetyandpharmacokinetics;antitumoractivityand biomarkers of response andresistancewerealso examined.Inaddition,wereportresultsfromanexpansioncohort intheongoingphase 1b partevaluatingsingle-agentD3S-001inpatients withNSCLCwho experienced disease progression after prior KRAS-G12Citherapy.

Results

Trial description

Theprimaryendpointswere assessmentofthesafetyandtolerability of D3S-001monotherapyin adult patients with KRAsGi2cmutant solid tumorsanddeterminationofthemaximumtolerateddose(MTD)and recommendedphase2dose(RP2D).Secondaryendpointswereassessmentofpharmacokinetics,ORR,durationofresponse(DOR),disease controlrate(DCR),DCRat24weeks andprogression-free survival (PFS)assessed per ResponseEvaluationCriteria inSolid Tumours,v.1.1, byinvestigatorsandblindedindependentcentralreview.Biomarkerswereassessedasexploratoryendpoints.Keyeligibilitycriteria wereage _{>=18} years,histologicallyorcytologicallyconfirmedlocally advancedormetastaticsolidtumors,documented K R A S^{G12C} mutation bylocaltestontumortissueorblood,measurablediseaseperResponse EvaluationCriteriainSolidTumours,v.1.1andEasternCooperative OncologyGroup(ECOG)performancestatusofOor1.Patientsinthe dose-escalationcohortsmusthavereceivedatleastonepriorlineof systemictherapyforadvanceddisease.PatientswithNSCLCinthe dose-expansioncohortswereeligibleiftheyhadreceivednomore thanonepriorKRAS-G12Cinhibitorandachieveddiseasecontrol. Afull list ofeligibility criteriais given inMethods.

Part1:phase la dose escalation

Patients.Asofthedatacutoffof16August2024,42patientswere enrolledinpart1,including25withNSCLC,13withCRCand4with PDAC(Table1).Thefirstpatientwasenrolledon17August2022,and thelastpatient was enrolledon21February2024.Overall,18patients (42.9%) hadreceived three ormore priorlines of systemictherapy, 7(16.7%) had brain metastases and 13 (31.0%) had livermetastases (Table1).Sixpatientshad received previousKRAS-G12Ci.Among the 13patientswithCRCenrolledinpart1,10 (76.9%) hadreceivedthreeor more prior lines of systemic therapy and7 (53.8%) hadlivermetastases (ExtendedData Table1).

Inpart1,patientsreceivedonce-dailyD3s-001at dosesof 50mg (n=3) ,100mg(n=7), 200mg(n=5) 400mg(n=9) 600mg(n=9) or 900mg\left(n=9\right) (Fig.1).Atthetimeof analysis,treatment had been discontinuedin26patients (61.9%) ;reasonsfordiscontinuation were: disease progression (n=18) ,death (n=4) ,patient withdrawal (n=2) orinvestigator decision (n=2) .Treatment wasongoingin16patients (38.1%) .ThemediandurationofD3S-001treatmentwas8.0months (range 0.2to 22.0 months).

Safety.Nodose-limitingtoxicities(DLTs)wereobservedoverthedose range studied,and the MTDwas not reached.Treatment-emergent adverseevents(AEs)ofanygradeoccurredin40ofthe42patients (95.2%) enrolledinpart1andwereconsideredrelatedto study treatmentin33patients (78.6%) (Table2).Grade3treatment-relatedAEs werereported inseven patients (16.7%) and included hypertriglyceridemia(onepatient),hyperkalemia(onepatient),nausea(one patient),abnormal hepaticfunction(one patient;gamma-glutamyl transferase(GGT)increased),increased lipase(twopatients),increased alanineaminotransferase(ALT)(twopatients),andincreasedaspartate aminotransferase(AST)(onepatient)(Extended DataTable2).Nograde 4or5treatment-relatedAEswerereported.Serioustreatment-related AEsoccurredinthreepatients (7.1%) .The mostfrequent 215% of patients)treatment-relatedAEsof anygradewere nausea( {\bf\zeta}_{n=19} 45.2%) ,diarrhea (n=13;31.0%) ,increasedamylase (n=8;19.0%) ,hypertriglyceridemia (n=8;19.0%) ,vomiting (n=8;19.0%) andincreased lipase (n=7;16.7%) (Table2).

Among25patientswithNSCLCenrolledinpart1,hepatotoxicity (definedasALTincreased,ASTincreased,GGTincreased orhepatic functionabnormal)occurredintwoofsix (33.3%) ,oneofseven (14.3%) andtwoofnine (22.2%) patientswhohadreceivedprior anti-programmed death1(PD-1)orprogrammed deathligand1(PD-L1) therapy \scriptstyle<=6 weeks,6-12 weeksor _{>=12} weeksbeforeinitiatingstudytreatmentwithD3S-001,respectively,andinoneofthree (33.3%) patients who had not received prior anti-PD-1orPD-L1therapy.Grade3GGT increaseoccurred in onepatient who hadreceivedprior anti-PD-1 orPD-L1therapy \scriptstyle<=6 weeksbeforeinitiatingD3S-001;all otherAEs of hepatotoxicityweregrade1inseverity.Hepatotoxicitywas not observed in thefour patients withNSCLCinthe 600*mg dose group, including onepatienteachwith \scriptstyle<=6 weeks,6-12 weeks and _{>=12} weeks ofprioranti-PD-L1treatment,andonepatientwithnoprioranti-PD-L1 treatment.Nointerstitiallungdiseaseorpneumonitiseventswere reportedinthedose-escalationphase.

Pharmacokinetics.Adose-dependentincreaseinD3S-001steadystateexposurewithminimalaccumulationwasobservedafterrepeated daily dosing.Atdoses >=400mg geometricmeansteady-stateconcentrationofdrugimmediatelybeforetheadministrationofnextdose (C_{trough}) (free drug) was >1 nM(Extended DataFig.1A),corresponding toexposuresrequiredfor 595% target inhibition in tumor cells.At doses >=400mg ,individualsteady-statearea underthecurve (\mathsf{A U C}_{tau}) (freedrug)was >33.6\mathsf{h}x\mathsf{n g m l}^{-1} (Extended Data Fig.1B),correspondingtoexposuresrequiredforcompletetumorregressioninmouse xenograft models.At the 600-mg dose,the C_{trough} (free drug)reached 1 nMand \mathsf{A U C}_{tau} (free drug)reached 33.6\mathsf{h}x\mathsf{n g m l}^{-1} for allpatients.The terminal t_{\nu_{2}} at steady state was approximately11h.An increaseinbioavailabilitywasobservedbetweenthe200-and400-mgdoses.Based onthese data,the600-mgdose wasselectedforfurtherinvestigation.

Antitumoractivity.Among42patientsenrolledinpart1,36were G12Ci-naive,34ofwhomwereincludedintheresponse-evaluableset (21NSCLC,9CRC,4PDAC).ConfirmedORRinthe34evaluablepatients was 73.5% d 95% confidence interval(CI) 55.6to 87.1)and the DCR was 97.1% 1%(95%{\bf C l}84.7 to99.9)(ExtendedDataTable3).ConfirmedORRwas 66.7% (95%0747.8 to88.7)among21evaluableG12Ci-naivepatientswith NSCLC and 88.9% 95% CI51.8 to 99.7) among9 G12Ci-naive patients withCRC;all4patients withPDAC demonstrated tumorshrinkage;3 confirmedpartialresponses(confirmedORR 75.0% 1 95% CI19.4to 99.4)). AhighORRwasobservedinpatientswithCRCinthisheavilypretreated population,themajorityofwhom had livermetastases(Extended Data Table1).Among G12Ci-naiveresponders,theestimated 6-month DOR ratewas 78.4% 1 95% CI55.6to90.4).Amongall36G12Ci-naivepatients, theestimated6-monthPFSratewas 68.6% 95% C150.5to81.3)(Extended DataTable3).AmongsixG12Ci-pretreatedpatients,fivewereincluded intheresponse-evaluableset,and allhadabestresponseofstable disease.Tumorshrinkagewasobservedacrossdoselevelsandtumor types (Fig.2a),and longitudinal change in tumor burden over time isshowninFig.2b-d.Durableresponseswereobservedacrosstumor types(Fig.2e).Amongsevenpatientswithbrainmetastasesatbaseline, twohadintracranialtumorshrinkageandfivehadstableintracranial disease.Representativescansofapatientwhoexperiencedintracranial tumorshrinkageareshowninExtendedDataFig.2.Themediandurationoftreatment without intracranialprogressionwas9.2months.

Biomarkeranalysis.Evaluationofthemolecularresponsedemonstrated that among34 evaluableG12Ci-naivepatientsinpart1,25 (73.5%) were K R A S^{G12C} circulating tumor DNA(ctDNA)positive at baseline(14NSCLC,7CRC,4PDAC).All25patientshadrapidreduction of K R A S^{c12C} mutant allele frequency (MAF) as early as cycle1, day 8 (Fig.3a).Thebestreductionin MAFwas {>}90% for23patients,including 20 who had 100% reduction.Confirmed ORR was 80.0% 95% CI 59.3to93.2)inpatientswhowereG12C-positiveatbaseline (n=25) comparedwith 55.6% 1 95% CI21.2to86.3)inpatientswhowereG12C negative at baseline (n=9) .AmongpatientswhowereG12C-positive atbaseline,confirmedORRwas 90.0% 1 95% CI68.3to98.8)in those who achieved complete K R A S^{G12C} MAFclearance (n=20) and 40.0% 95% CI5.3to85.3)inthosewhodidnot (n=5) .Amongthe14patients positivefor K R A S^{c12C} ctDNAatbaseline,6hadSTK1lco-mutations,allof whomachievedconfirmedresponse. K R A S^{G12X} orKRASswitchllpocket co-mutationsthatmayconferprimaryresistancetoG12Ciwerenot observedinthiscohort.OnepatientwithCRCwhodidnotrespond toD3S-001treatment(progressivedisease)hadhighMAFAPC (62%) and T P S3(68%) co-mutations inadditionto K R A S^{c12C} (40%) .Inaddition, amongfive evaluableG12Ci-pretreatedpatients,four (80.0%) were K R A S^{{GI2C}} ctDNA positive at baseline (one NSCLC,three CRC).All four patients had reduction of K R A S^{G12C} MAF by cycle 1, day 8 (Fig. 3a),and allachievedstabledisease asthebestoverallresponse.

Fatl

AcquiredgenealterationsandKRASstatuswereassessedatend oftreatment(EOT)inllpatientswhohaddiseaseprogressionandhad EOTctDNAsamplescollected(Fig.3b).MutationprofilesatEOTwere heterogeneous andattributedtoreboundofpre-existingmutations detectedatbaseline(ExtendedDataTable4).Acquiredgenealterations that werenotpresent atbaselineweredetectedatEOTin5of11patients. No KRASG12c amplification was observed at EOT.Acquired secondary KRASmutationsandMAPK/RTKpathway-relatedgenealterationswere detected atEOTin two patientswith CRC. K R A S^{c12C} MAFvalues at EOT wereeitherlowerthanbaselineorundetectable(Fig.3c).

Part2:phase1bdoseexpansion

Patients.Asofthedatacutoffof13December2024,20patientswith KRAS-G12Ci-pretreatedNSCLCwere enrolled in the dose-expansion cohort and received D3s-001at the selected phase 2 dose of 600{mg} once daily(Table1).Thefirst patient was enrolledon11January 2024, andthelastpatientwasenrolledon2August2024.PreviousG12Ci treatmentwassotorasib(sevenpatients),garsorasib(fourpatients), fulzerasib(twopatients),glecirasib(two patients),adagrasib(one patient),BPI-421286(one patient),HS-10370 (one patient),JDQ-443 (one patient)orolmorasib(onepatient).BestresponsetopriorG12Ci treatmentwaspartialresponseineightpatients (40.0%) ,stabledisease insevenpatients (35.0%) and unknown infivepatients (25.0%) .Prior G12Citreatment wasdiscontinuedbecause ofprogressivedisease in19patients (95.0%) andbecauseoftoxicityin1patient (5.0%) ;19 patients (95.0% hadreceivedpriorG12Citreatmentfor >=6 monthsand 14 patients (70.0%) wereenrolledimmediatelyafterdiseaseprogressionfollowingpriorG12Citreatment(ExtendedDataTable5).

Atthetimeofanalysis,treatmenthadbeendiscontinuedin 1lpatients (55.0%) ;reasonsfordiscontinuationwere diseaseprogression(9patients),patientwithdrawal(1patient)orinvestigatordecision (1patient).Treatmentwasongoinginninepatients (45.0%)

Safety.Thesafetyprofilewasconsistentwiththatobservedwiththe 600-mgdoseinpart1.Treatment-emergentAEsofanygradeoccurred in18ofthe20patients (90.0%) enrolledinthedose-expansioncohort andallwereconsideredrelatedtostudytreatment.Grade3orhigher treatment-relatedAEsoccurredintwopatients (10.0%) andserious treatment-relatedAEsoccurredintwopatients (10.0%) (Extended Data Table 6).

Amongthe20patientswithKRAS-G12Ci-pretreatedNSCLC enrolled in part 2,16 (80.0%) had received anti-PD-1orPD-L1therapy >=12 weeksbeforeinitiatingD3S-001studytreatment.Treatment-related hepatotoxicity(definedasALTincreased,ASTincreased,GGTincreased orhepaticfunctionabnormal)occurredintwopatients (10.0%) ;both weregrade1events.Oneofthesepatientshadnotreceivedprior anti-PD-1orPD-L1therapyandtheotherhadreceivedanti-PD-1or PD-L1therapy {>=}12 weeksbeforeinitiatingD3S-001.Nointerstitiallung disease orpneumonitis wasreportedinthiscohort.

Pharmacokinetics.Steady-state pharmacokinetics ofD3s-001 wereconsistentwiththoseobservedwiththe 600*mg doseinpart1 (Extended DataFig.3).At steady state,mean ± s.d.maximum plasma concentration (C_{max}) was 507±275{ng{ml^{-1}}} and \mathsf{A U C}_{\mathsf{t a u}} was ^{4,740±} 2,510\mathsf{h}x\mathsf{n g}\mathsf{m l}^{-1} .The terminal t_{\nu_{2}} at steadystatewas 7.93±1.35\mathsf{h}

Antitumoractivity.Tumorshrinkagewasobservedin12ofthe 20 patients (60.0%) in the dose-expansion cohort(Fig.4a).Partial responsewasachievedinsixpatients (30.0%) ,with responses confirmedinfourpatients (20.0%) ,andtheDCRwas 80.0% Longitudinal changeintumorburdenovertimeisshowninFig.4b,andtreatment durationand response perpatient is shown in Fig.4c.

Biomarkeranalysis.Evaluationofmolecularresponsedemonstratedthatamong20KRAS-G12Ci-pretreatedNSCLCpatientsin the dose-expansion cohort,14 (70.0%) were K R A S^{G12C} ctDNA positive atbaseline.Baselineco-mutationsincludedsecondaryKRAS alterations(A146V,R68SandG12Camplification)andBRAF,EGFR, NTRK and TP53 alterations.The best reduction in K R A S^{{G12C}} MAF was {>}90% for11 patients,including 8who had 100% reduction.All six radiologicresponders(partialresponse)were K R A S^{{G12C}} ctDNA positive at baseline andfive of thesepatients had completeclearance of K R A S^{{G12C}} MAF.

AmongthreepatientswithKRAsGi2camplificationatbaseline,two achievedpartialresponseandoneachievedstabledisease(Extended DataFig.4).Conversely,patients whohadimmediatediseaseprogression onD3S-001hadbaseline gene alterations of K R A S^{R689} B R A F^{\mathsf{V600E}} ， MYCamplification,NTRK2,CDKN2AandPIK3CG.

Discussion

First-generationGDP-boundKRAS-G12Ci,suchassotorasibandadagrasib,havedemonstratedclinicalproofofconceptbutarelimitedin depthand durationof responses13-162324 OtherGDP-bound KRAS-G12Ci inclinicaldevelopmentincludeglecirasib5,olomorasib26anddivara\sin^{27} .An alternative strategyinvolves targeting GTP-bound KRAS,as exemplified by KRAS-G12C(ON)inhibitors such as RMC-6291;early clinicaldatafromphase1trialsdemonstratedpromisingactivityforthis approach28.29Sufficienttarget coverageisnecessary withanyapproach.

WereportclinicaldataofD3S-001,anext-generationGDP-bound KRAS-G12Ci.Preclinically,D3S-001demonstratedseveralkeyfeaturesthathighlightitspotentialtoovercomethelimitationsof first-generationinhibitors19.Itexhibits exceptionally highpotency, achievingcompletetargetinhibitionat1nM,aconcentrationattainable at clinically relevant C_{trough} concentrations".Furthermore,it demonstratesrapidtargetengagementkinetics,withareactionspeedsurpassingthatofKRASnucleotideexchange,effectivelydepletingactiveKRAS eveninthepresenceofgrowthfactors9.Thesefeaturesarepredictedto enhance single-agent efficacy in K R A S^{{G12C}} -driventumors,particularly CRCwherehighRTKactivationisprevalentandEGFR-mediatedKRAS reactivationisamajorresistance mechanism20-22.Inaddition,D3S-001 hasdemonstratedhighselectivity,being >2,000 and {>}5,000 foldmore potentin K R A S^{c12c} mutantcompared with non-G12C KRASmutant cell linesintwo-andthree-dimensionalproliferationstudies,respectively19. Asacovalentinhibitor,D3S-001showednodetectablenonspecificfree cysteineconjugationinwhole-proteomeprofilingassays.Based on its high selectivity and the wide therapeutic index predicted during toxicologystudies(predictedefficaciousexposure17-and9-foldbelow thepharmacokineticexposureatthehighest dosethat was notseverely toxicbasedon C_{max} andAuC,respectively,nagoodlaboratorypractice studyindogs),afavorablesafetyprofilewas anticipatedin humans30.

In the dose-escalationpart ofthis study (part 1),D3s-001 demonstratedafavorablesafetyprofile,andclinicalpharmacokinetic exposurereachedthedesiredthresholdforcompletetargetinhibition (free drug C_{trough}{>}1{\mathsf{n M}} .At doses {>}400~mg all patients achievedfree drugplasmaconcentrationsassociated withcompletetargetengagement and tumorregressioninpreclinical studies19.Aonce-daily dose of 600mg wasselectedasapotentialphase2dose.Thesafetyprofile of D3s-001wasgenerallyconsistent withthose of first-generation KRAS-G12Cinhibitors13-16.23.24Theseresults supportfurtherevaluation ofD3s-o01asmonotherapyandincombinationwithotheranticancer therapies.

Fig.2|AntitumoractivityofD3S-001inpart1(phaseladoseescalation). a,Waterfall plotofmaximum percentagechangefrombaselineinsumof diametersoftargetlesions inindividualpatients.The asteriskindicates patients withongoingtreatment.b-d,Spiderplotsofthepercentagechangefrombaseline insumofdiametersovertimeinindividualpatients withNSCLC(b),CRC(c)and PDAC(d).e,Swimlaneplotoftreatmentdurationandresponseinindividual

patients.Purplediamondsrepresentfirstdocumentedresponse,andredtriangles represent progressivedisease.The waterfall(a)andspider(b-d)plots were based ontheresponse-evaluableset,whichincludedallpatientswhohadatleastone post-baselineimage.Twopatients whodiscontinuedstudytreatmentwithoutany post-baselineimagearerepresentedasnotevaluated.C,CRC;N,NSCLC;P,PDAC; PD,progressive disease;PR,partialresponse;SD,stable disease.

Inpart1ofthisstudy,promisingantitumoractivitywasobserved acrosstumortypesinKRAS-G12Ci-naivepatients.Theoverall confirmedORRwas 73.5% withaDCRof 97.1% .ConfirmedORRsin patientswithNSCLC,CRCand PDACwere 66.7% 88.9% and 75.0% respectively.D3s-001thereforedemonstratedhigherresponserates thanfirst-generationinhibitors.Amongthe21patientswithG12Ci-naive NSCLCfromthe dose-escalationcohortofthis study,theconfirmed ORRwaS 66.7% withD3S-001treatment,whereasfirst-generationG12C inhibitorshavedemonstratedORRsrangingfrom 28% to 43% (refs. 13,14,23,24).Ourdata arerelativelyimmatureforformalassessmentof PFS.However,the6-month DORand PFSrates in patients with NSCLC were 77.4% and 66.5% respectively;therefore,themedianDORorPFS mustbelongerthan6months.IntheCodeBreak200andKRYSTAL-12 studies,themedianPFSwithsotorasiband adagrasibwas5.6and 5.5months,respectively1314.31.The novelKRAS-G12Ci divarasib also reportedahigherresponserate andbetterdurabilityinpatientswith NSCLCrelativetofirst-generationG12Ci,withanORRof 53.4% and medianPFSof13.1monthsin60patientswithoutpreviousexposure toG12Ciinaphase1study27.TheclinicalefficacyofD3S-001inpatients with NSCLCappearstobebetterthanthefirst-generationKRAS-G12Ci and comparablewithdivarasib.ThecombinationofD3S-001at 600{mg} oncedailywithpembrolizumabisbeingassessedasfirst-linetreatment for K R A S^{G12C} mutant NSCLC in a dose-expansion cohort.

Fig.3|Molecularresponseinpart1(phaseladose escalation).a,Percentage change inKRASG2cMAFinctDNAbytumortype andbestresponse.b,Acquired gene alterations and KRAS status at EOT.c, K R A S^{{G12C}} MAF dynamic in patients withevaluablectDNAEOTsamples.Panelsbandcincludeonepatientforwhom thesamplewasobtainedbeforethedatacutoffdateof9September2024,but forwhomdatafromsampleanalysiswerenotavailableuntilafterthedatacutoff
date.c,OrangelinesrepresentpatientswithNsCLCand bluelinesrepresent patients with CRC.Patient IDis shown atthe corresponding EOTtimepoint.BOR, best overall response;C4D1,day1of cycle4;CNV,copy numbervariation;Indel, insertionordeletion;N.A.,notapplicable;N.D.,not detected;Neg,negative; PI3K,phosphoinositide3-kinase;Pos,positive;RAS,rat sarcoma;SNV,single nucleotidevariation.

In the ninepatients withCRCfromthedose-escalationcohort of this study,a high confirmed response rate (ORR 88.9% and prolonged response duration(6-monthDOR, 85.7% wereobserved. These outcomessurpassed ORRsobservedwithfirst-generation KRAS-G12Cicombined with anti-EGFR antibodiesinpatients with CRC, which ranged from 26% to 46%^{16-19} Although the sample size islimited,thesedataprovideearlyclinicalevidencethatunderscores theimportanceofrapid andefficienttargetengagementinovercoming theEGF-stimulatedGDP-to-GTPtransitioninCRC.Expansioncohorts areplannedtovalidatethesefindings.Currentresultssupportfurther validationofD3S-o01asmonotherapyinCRCandexplorationof combinationstrategieswithanti-EGFRantibodiestomitigateMAPK pathwayreactivationmediated bywild-typeKRAS,NRASand HRAS alleles.Preclinically,althoughsingle-agentD3s-001treatmentresulted inrobustantitumorresponsesin 40perthousand ofpatient-derivedxenograft models (n=10) ,combinationwithcetuximabfurtherenhancedthe depthofresponse,with 80% ofpatient-derivedxenograftmodels demonstrating >30% tumorregression approximately21days after treatment'9.Thishighlights the potentialforbypass activation of non {*}K R A S^{G12C} genesregulatedbyGFRactivationinRC,withcertan genetic compositionspotentiallyfavoringthistype of bypass activation more than others.

Inanexpansioncohortof20NsCLCpatientswhosediseasehad progressedonfirst-generationKRAS-G12Ci(part2),D3S-001demonstratedanORRof 30.0% andaDCRof 80.0% .Thisstudyrepresents clinical datareportingsingle-agent activityofaGDP-boundKRAS inhibitorinafocusedcohortofpatientswithNsCLCwhohaddisease progression on G12Ci.This clinical benefit is akeyfinding consideringthatpatientswerepreviouslytreatedwithvariousGi2Ciandhad diverseresistancemechanisms,andprovidesearly clinicalvalidation of themechanismof actionof D3S-o01translatingintoenhanced activityoverfirst-generationG12Ci.Despiteahighmutationload and heterogeneous mutation landscape atbaseline,the ctDNA dynamics demonstratedsubstantialmolecularinhibitionoftheKRAsGi2c-positive clonebyD3s-001eveninthepresenceofco-mutations associated withpoorprognosis,includingEGFR,NTRK1andCDKN2A.Conversely, genealterations ofKRASswitchll pocket and B R A F^{\mathsf{V600E}} likely acquired frompreviousG12Citreatment,mayhavecontributedtotheprimary resistancetoD3S-001(ref.19).Amongthe20patients,3hadKRASgene amplificationsatbaseline,aknownresistancemechanismtoearlier inhibitors3233;2ofthesepatients achievedapartialresponse withD3S001andtheremainingpatienthadstabledisease.Bothresponderswith KRASgene amplifications received D3S-001asthe next line oftherapy immediately afterprogressing onthefirstG12Ci.The antitumoreffect withthismechanism ofresistance aligns with preclinicaldatashowing activityina K R A S^{c12C} gene amplification xenograft model9.

ThemolecularprofileatEOTsuggeststhatmechanismsofresistanceforD3S-001aresimilartootherKRAS-G12Cisotorasib,adagrasib and divarasib273233 Twopatients withCRC acquiredsecondary K R A S^{c12X} alterations(G12D,G12V),secondaryKRASalterationsassociatedwith theswitchIIpocket(Q61H,H95N)andMAPK/RTKpathway-related gene alterationsfollowingD3S-001treatmentin this study.Unlike the other KRAS-G12C inhibitors,in which K R A S^{G12C} amplification has been reported in 10% to 15% of patients with disease progression on adagrasib or sotorasib32.3 and 56% of those with disease progression on divarasib27, no K R A S^{G12C} amplifications were observed inthe small numberofsamplesfrompatientswithNsCLCtestedtodateinthisstudy usingsimilardetectionmethods(ctDNAanalysisusingnext-generation sequencingon liquid biopsy at EOT).The lack of acquired K R A S^{c12C} amplificationsfollowingD3S-001andclinicalefficacyinpatientscarrying K R A S^{G12C} amplificationsatbaselinesuggestthatcompletetarget engagementofD3s-001maypreventthisbypassresistancemechanism.

Althoughthispreliminary clinicalevidenceforthesafetyand antitumor activity of D3s-001is encouraging,this study was limited bythesmallsamplesizeofpatientsenrolledintoparts1and2and lackofcomparison withastandard-of-caretreatmentinpatientswith K R A S^{G12C} mutation.

Inconclusion,thisstudyrepresentstheclinicaltranslationofthe next-generationGDP-boundKRAS-G12Ci,D3S-001andhighlights the potentialofD3S-001toovercomekeylimitationsofearlierG12Ci. Strategies aimed at enhancingclinical outcomesthroughimproving compoundpotencyandtargetcoveragehavebeencrucialinadvancing small-moleculetargetedtherapies.InthecontextofKRAS-G12Ci,divarasib,whichis5-20times morepotentthansotorasib andadagrasib34, hasdemonstratedhigheroverallresponseratesandlongerdurations ofresponse.ThisstudyintroducesD3s-001,whichexhibitstwoorders ofmagnitudegreaterpotencyandtargetengagementefficiencycompared withfirst-generation G12Ciand represents a next-generation inhibitorwithpotentialtobecomethebackboneoftreatmentfor K R A S^{{G12C}} mutated solid tumors.Clinicalinvestigations are ongoingin K R A S^{{G12C}} mutated solid tumors across various disease settings and in largerpopulationstovalidatethepotentialof D3S-o01inimproving clinicaloutcomesinthisdiseasearea,where anunmetmedical need persists.

Online content

Anymethods,additionalreferences,NaturePortfolioreportingsummaries,sourcedata,extendeddata,supplementaryinformation, acknowledgements,peerreviewinformation;detailsofauthorcontributionsandcompetinginterests;andstatementsofdata andcodeavailabilityare available athttps://doi.org/10.1038/s41591-025-03688-6.

References

1 Zehir,A.et al.Mutational landscape of metastatic cancer revealedfromprospectiveclinicalsequencingof10,oo0patients. Nat.Med. 23, 703-713 (2017).

I1utduUs I1al11. dle1s.1, 24/-40/ (U1) 3. Nassar, A.H., Adib, E.& Kwiatkowski, D. J. Distribution of KRAS (G12C)somaticmutations acrossrace,sex,and cancertype. N.Engl. J. Med. 384, 185-187(2021).
4.Frost, M.G.et al.KRAS G12C mutated advanced non-small cell lung cancer (NSCLC): characteristics, treatment patterns and overall survival from a Danish nationwide observational register study.Lung Cancer 178,172-182(2023).
5. Sebastian, M.et al.KRAS G12C-mutated advanced non-small cell lung cancer:a real-world cohort from the German prospective, observational,nation-wide CRISP Registry(AIO-TRK-0315).Lung Cancer154,51-61(2021).
6. Salem,M.et al. O-3 characterization of KRAS mutation variants andprevalenceofKRAS-G12Cingastrointestinal malignancies. Ann. Oncol. 32, S218 (2021).
7. Kolch, W.,Berta, D.&Rosta,E.Dynamic regulation of RAS and RAS signaling.Biochem.J. 480,1-23 (2023).
8. Huang, L., Guo, Z.,Wang, F. & Fu, L. KRAS mutation: from undruggable to druggable in cancer.Signal Transduct.Target. Ther.6,386 (2021). 9. Hunter,J. C.et al.Biochemical and structural analysis of common cancer-associated KRAS mutations.Mol. Cancer Res.13, 1325-1335 (2015).
10. Patricelli, M.P.et al. Selective inhibition of oncogenic KRAS output with small molecules targeting the inactive state. Cancer Discov. 6, 316-329 (2016).
11.Ostrem, J. M., Peters, U.,Sos, M. L., Wells, J. A.&Shokat, K.M. K-Ras(G12C) inhibitors allosterically control GTP affinity and effector interactions.Nature 503,548-551(2013).
12.Janes, M.R.et al. Targeting KRAS mutant cancers with a covalent G12C-specific inhibitor. Cell 172,578-589.e517 (2018).
13. Mok, T.S.K.et al. KRYSTAL-12: phase 3 study of adagrasib versus docetaxel in patients with previously treated advanced/ metastatic non-small cell lung cancer(NSCLC) harboring a KRASG12C mutation.J. Clin.Oncol. 42, LBA8509 (2024).
14. de Langen, A.J.et al. Sotorasib versus docetaxel for previously treated non-small-cell lung cancer with KRAS(G12C) mutation: a randomised,open-label,phase 3 trial.Lancet 401,733-746 (2023).
15.Fakih, M.G.et al. Sotorasib for previously treated colorectal cancers with KRAS(G12C) mutation(CodeBreaK100): a prespecified analysis of a single-arm, phase 2 trial. Lancet Oncol. 23,115-124 (2022).
16.Yaeger,R.et al.Adagrasib with or without cetuximab in colorectal cancer with mutated KRAS G12C.N.Engl. J. Med.388,44-54 (2023).
17.Fakih, M.G.et al. Sotorasib plus panitumumab in refractory colorectal cancer with mutated KRAS G12C.N.Engl.J.Med.389, 2125-2139 (2023).
18.Yaeger, R.et al. fficacy and safety of adagrasib plus cetuximab in patients with KRASG12C-mutated metastatic colorectal cancer. CancerDiscov.14,982-993 (2024).
19. Zhang,J.et al. D3S-001, a KRAS G12C inhibitor with rapid target engagement kinetics, overcomes nucleotide cycling,and demonstratesrobustpreclinicalandclinicalactivities.Cancer Discov.14,1675-1698 (2024).
20. Ryan, M.B.et al. KRAS(G12C)-independent feedback activation of wild-type RAS constrains KRAS(G12C) inhibitor efficacy. Cell Rep. 39, 110993 (2022).
21.Xue, J. Y.et al. Rapid non-uniform adaptation to conformationspecific KRAS(G12C) inhibition.Nature 577,421-425 (2020).
22. Amodio, V. et al. EGFR blockade reverts resistance to KRAS(G12C) inhibition in colorectal cancer.Cancer Discov.10,1129-1139 (2020).
23.Janne,P. A.et al.Adagrasib in non-small-cell lung cancer harboring aKRAS(G12C)mutation.N.Engl.J.Med.387,120-131(2022).
24. Skoulidis,F.et al. Sotorasib for lung cancers with KRAS p.G12C mutation.N.Engl.J.Med.384,2371-2381(2021).
25. Shi, Y. et al. Glecirasib in KRAS(G12C)-mutated nonsmall-cell lung cancer:aphase 2b trial.Nat.Med.31,894-900(2025).
26.Burns, T.F.et al.Efficacy and safety of olomorasib (LY3537982),a second-generation KRAS G12C inhibitor(G12Ci), in combination withpembrolizumabinpatientswithKRASG12C-mutant advanced NSCLC.J.Clin.Oncol.42(Suppl.16),abstr8510(2024).
27. Sacher, A.et al. Single-agent divarasib (GDC-6036) in solid tumors withaKRAS G12Cmutation.N.Engl.J.Med.389,710-721(2023).
28.Janne,P.A.et al.AbstractPRO14:preliminary safety and anti-tumor activity ofRMC-6291,afirst-in-class,tri-complexKRASG12C(ON) inhibitor,inpatientswithorwithoutpriorKRASG12C(OFF)inhibitor treatment.Mol.CancerTher.22,PR014(2023).
29.Nokin, M.J.et al.RAS-ON inhibition overcomes clinical resistance toKRASG12C-OFFcovalentblockade.Nat.Commun.15,7554 (2024).
30.Lu,J.etal.D3S-001,a highly potent,selective,and differentiated covalentinhibitorofKRASG12C:humandoseprediction and first-in-human(FIH) trial design.J.Clin.Oncol.40,e15087(2022).
31.Lee, A.T. M.& Nagasaka, M.Adagrasib in KRYSTAL-12 has not brokentheKRASG12Cenigmacodeoftheunspoken6-month PFS barrier in NSCLC.Lung Cancer(Auckl.)15, 169-176 (2024).
32.Awad,M.M.et al.Acquired resistance to KRAS(G12C) inhibition in cancer.N.Engl.J.Med.384,2382-2393(2021).
33.Zhao,Y.etal.Diverse alterations associatedwithresistanceto KRAS(G12C) inhibition.Nature 599,679-683(2021).
34.Brazel, D.& Nagasaka,M.Divarasib in the evolving landscape ofKRASG12CinhibitorsforNSCLC.Target.Oncol.19,297-301 (2024).

Publisher's note Springer Nature remains neutral with regard to jurisdictionalclaimsinpublished mapsand institutional affiliations.

SpringerNature oritslicensor(e.g.a societyorotherpartner)holds exclusiverightstothis articleunderapublishingagreementwith the author(s)orotherrightsholder(s);authorself-archivingofthe acceptedmanuscriptversionofthisarticleissolelygovernedbythe terms of such publishing agreement and applicable law.

⊚ The Author(s),under exclusive licence to Springer Nature America, Inc.2025

Methods

Study design and treatment

Thisisafirst-in-human,multicenter,open-label,phase1and2trial being conducted in three parts,comprising D3S-001 monotherapy dose escalation(part1),D3s-001monotherapy expansioncohorts (part 2)and D3S-001combinationtherapyexpansioncohorts (part 3). D3S-001was administeredorally once dailyin21-daytreatmentcycles, andtreatmentwascontinueduntildiseaseprogression,unmanageable toxicityorpatientwithdrawal.Thestudyisongoing,andpatientswere stillreceivingtreatmentatthetimeofpublication.

Inpart1,doseescalationwasperformedusingaBayesianoptimal intervaldesignwitha _{3+3} run-in,whichwasusedtodeterminetheMTD, tominimizetheprobabilityof incorrectdose assignment andtoguide doseescalation or de-escalation.This approachminimized therisk of exposingpatientstodoses abovetheMTD and hadahighprobability foridentifyingthecorrectMTD.Thestarting doseof D3S-001was50 mg oncedailyincohort1,withplanneddoseescalationupto9o0mgonce dailyoverfiveadditionalcohorts.DoselevelsweredeterminedfollowingamodifiedFibonaccimethod.Decisionstoexpandthenumberof patients atagiven doselevel,proceed tothe next doselevel,stop dose escalationorde-escalatetoalower dose levelwere madebyasafety reviewcommittee.Doseescalationwas allowedforpatientswhocompletedtwocyclesoftreatmentattheirassigneddoselevelwithoutexperiencingaDLTduringcycle1oranygrade 22 treatment-relatedAE;patients whometthesecriteriacouldproceedtothenextdoselevel,ifdetermined assafebythesafetyreviewcommittee,forthefollowingcycle.

Part 2isongoing and exploresD3s-001 monotherapyat the RP2D determined duringpart1in larger cohortsofpatients with previouslytreatedNSCLC,CRCandPDAC.Resultsforthe doseexpansioncohort,exploringD3S-001monotherapyinpatients withKRAS-G12Cinhibitor-pretreatedNSCLC arereported here; resultsforotherexpansioncohortswillbereportedseparately.

Part3isongoingandwillexploreD3S-001attheRP2Daloneorin combination with pembrolizumab or platinum doublet chemotherapy infirst-linetreatmentofNSCLCandD3S-001aloneorincombination withcetuximabinsecond-orlater-linetreatmentofCRC.

Trialoversight

Theinstitutional reviewboard ateachstudy site(Supplementary Table1)approvedthetrialprotocol.ThetrialwasconductedinaccordancewiththeGoodClinicalPracticeguidelinesoftheInternational CouncilforHarmonisationofTechnicalRequirementsforPharmaceuticalsfor Human Use,theprinciplesoftheDeclarationofHelsinki, andlocalregulationsregardingtheconductofclinicalresearch.All patientsprovidedwritteninformedconsentbeforeparticipatinginthe trial.Safetyoversightwasconductedbythesafetyreviewcommittee,whichreviewedthesafetyprofileduringthedose-escalation phase(part1)andisinvolvedinongoingsafetyreviewduringpart2. Allauthorsensuredthecompletenessandaccuracyofthedataand analyses,andthefidelityofthetrialtotheprotocol.Protocolamendments arelistedinSupplementaryTable2.

Patients

Eligiblepatientswereaged≥18years,hadalifeexpectancyof >=12 weeks accordingtotheinvestigator,were abletocomprehend and willing tosignaninformedconsentform,hadhistologicallyorcytologically confirmed locallyadvancedormetastaticsolidtumors,documented K R A S^{G12C} mutationin the past 5 years by localtest ontumor tissue or blood,measurablediseaseperResponseEvaluation CriteriainSolid Tumoursv.1.1,andanECoGperformancestatusof0or1,and must havehadadequateorganand marrowfunctioninthescreeningperiod. Patientsmusthavebeenabletotakeoralmedication,eatastandardized high-fat,high-calorie meal within25min andbe abletofastfor \mathtt{\ge10\ h} ： For dose-escalation cohorts,eligible patients must have received at least one prior line of standard-of-caresystemic therapyforlocally advancedandunresectableormetastaticdisease.Inthefirstthreedose cohorts(50,100 and 200mg patientswithpriorKRAS-G12Cinhibitor treatmentwereexcluded.Indosecohorts >=400mg patientswhohad receivednomorethanonepriorKRAS-G12Cinhibitorwereallowedif theyhadachieveddiseasecontrol(completeorpartialresponse or stablediseasefor ^{>=6} months)andhadnotpermanentlydiscontinued treatmentbecauseoftreatment-relatedAEs.Thenumberofpatients withpriorKRAS-G12Cinhibitortreatmentwascappedat nomorethan one-thirdofthepatientsateachdoselevel.

For dose expansion in patients withKRAS-G12C inhibitorpretreatedNsCLC,patientswereeligibleiftheyhadreceivednomore thanonepriorKRAS-G12Cinhibitorand had achieveddiseasecontrol;patientswhodiscontinuedtreatmentwithin6monthscouldbe includedifdiscontinuationwasbecauseofsafetyortolerabilityreasons only.Patientswereexcludediftheyhadknownnon-G12CKRASmutations,EGFR-sensitizingmutations,ALK/ROS1/RETrearrangement, NTRK1/2/3 gene fusion, B R A F^{\mathsf{V600E}} mutation or METexon 14 skipping mutation.Patientswithtreatedoruntreatedbrainmetastaseswere eligibleforinclusioniftheyhad undergoneresectionorradiotherapy completedatleast14daysbeforeinitiationofstudytreatmentwith noevidenceofprogressionornewlesionsonfollow-upmagnetic resonanceimaging(MRi).Patientswereexcludediftheyhaduntreated and/oruncontrolledbrainmetastases,hadresidualneurologicsymptomsofgrade ^{>2} orwerereceivingcorticosteroids >=10mg perday prednisone orequivalent.

Otherexclusioncriteriaweremixedsmallcelllungcancerand non-small-cell histologyforNSCLCexpansion cohorts,uncontrolled intercurrentillness,current interstitiallung disease orpneumonitis or a previoushistoryofinterstitiallungdiseaseorpneumonitisrequiring high-doseglucocorticoids,oractiveleptomeningealdisease,spinal cordcompression,uncontrolledpleuraleffusion,pericardialeffusion orascitesrequiringmorethanoneplacementofacatheterpermonth. Patientscouldnothaveunresolvedtreatment-relatedgrade 22 toxic eventsfrompreviousanticancertreatment(exceptvitiligooralopecia). KRAS-G12Ci-pretreatedpatientscould nothave grade \scriptstyle>=3 ALT and/or ASTelevationat more thantwooccurrences orgrade 3orhighercardiactoxicity.Patientscouldnothave activegastrointestinaldiseaseor otherconditionsthatcouldinterferewiththeabsorption,distribution, metabolismorexcretionoforaltherapy.Nohistoryofallogeneicorgan transplantationwasallowed,andpatientsmustnothavehadinfection requiringtreatment withsystemicantibacterial,antifungalorantiviral therapywithin14daysbeforereceivingthefirstdoseofstudytreatment.Patients could not have untreated chronic hepatitis Bor bea carrier,haveactivehepatitisCorhaveahistoryofhumanimmunodeficiency virusinfection.Patientscould nothavea historyofothermalignancyexceptformalignancytreatedwithcurativeintentandnoknown activediseaseforatleast2yearsandlowriskofrecurrence,adequately treatednonmelanoma skincancerorlentigomaligna withoutevidence ofdisease,adequatelytreatedcarcinomainsituwithoutevidenceof disease,orlocalizednoninvasiveprimarydiseasethatdoesnotrequire treatment orimpact lifeexpectancy.Patients wereexcludedif they had corrected QTinterval >470ms (females)or >450 ms(males), increased riskofQTinterval prolongations orarrythmias,clinically importantelectrocardiogram abnormalitiesoruntreated sinusnode dysfunctionorothercardiovasculardisease.Patientscouldnothave active orpreviouslydocumented autoimmune orinflammatorydisorders andmust nothaveCOvID-19infection.Patientshadtobeable tostoppriororconcomitanttherapyforatleast14daysbeforereceivingthefirstdoseofstudytreatment.Patientsmustnothaveknown hypersensitivityto the studyinterventionorexcipients andcould not participateconcurrentlyinanyinvestigationalclinicalresearch.

Assessments

Theprimaryobjectivesweretoevaluatethesafety andtolerability of D3S-001and to determine the MTD and RP2D.Safety and tolerability wereassessedbytreatment-emergentAEs,treatment-relatedAEsand clinicallyrelevantchangesinvitalsigns,physicalexams,electrocardiogramsandclinicallaboratorytests.AEsweregradedaccordingto NationalCancerInstituteCommonTerminologyCriteriaforAdverse Eventsv.5.o.DLTs weredefined asanyofthefollowingevents occurring duringthefirstcycleoftreatmentinthedose-escalationstage:febrile neutropenia;neutropenicinfection;grade4neutropenia;grade3 thrombocytopeniawithbleeding;grade4thrombocytopenia;grade 4anemia;grade {>=}4 vomitingordiarrhea;grade ^{\ge3} nauseafor ^{\ge3} days despite optimalsupport;ALTorASTmorethanthreetimestheupper limitofnormalwithconcurrenttotalbilirubinmorethantwotimes theupperlimitofnormalwithoutevidenceofcholestasisoralternative explanations;any death not clearly dueto underlying disease or extraneouscauses;oranyothergrade \scriptstyle>=3 AEexcepttransientisolated laboratoryabnormalitiesthatresolvedtobaselinewithin72hwithout intervention,alopecia,grade3fatiguelasting ^{<7} days,grade3electrolyteabnormalitiesthatresolvedtograde <=1 within 48\mathsf{h} grade3rash thatresolvedtograde \scriptstyle<=2 within7dayswithmedicalmanagement,or grade3or4 asymptomaticelevationinserumamylase orlipase that wasnotassociatedwithclinicalorradiologicevidenceofpancreatitis. TheMTDwasdefinedasthedoselevelforwhichtheestimatedprobabilityofaDLTwasclosestto 25% usinganisotonicregressionmodel. TheRP2DwasdeterminedbasedontheMTD,pharmacokineticand pharmacodynamicinformation.

Secondaryobjectivesweretocharacterizethepharmacokinetics ofD3S-o01andtoevaluateantitumoractivity.Pharmacokinetic parametersincluding C_{max}, timetomaximumplasmaconcentration ({t_{\operatorname*{max}}}) ， t_{\nu_{2}} and AuCwere determined using standard noncompartmentalmethods.Pharmacokineticsampleswerecollectedwithin 30 min predose,and at0.5,1,2,4,6,12 and 24\mathsf{h} post dose on day1 ofcycle1andonday1ofcycle2.Additionalsampleswerecollected within 30min predose onday8of cycle1,and within 30{min} predoseandat 4\mathfrak{h} postdose onday1ofcycles3,5,7and9.Antitumor activity was measured by ORR(complete or partial response,confirmedbyrepeat assessment {>=}4 weeks afterinitial documentation), DCR(completeorpartialresponseorstablediseasemaintained for \scriptstyle>=24 weeks),DoR(timefromfirstdocumentationofresponse until diseaseprogression or deathfrom anycause)andPFS(time fromfirstdoseofstudytreatmentuntildiseaseprogressionordeath fromanycause)perResponseEvaluationCriteriainSolidTumours v.1.1.Tumorassessmentswere performed bycomputedtomography orMRIevery6weeksuntilweek25,thenevery12weeksthereafter. Patientswithoutpost-baselinediseaseassessmentswereconsidered nonresponders.

Exploratoryobjectivesweretoinvestigatepotentialbiomarkersby biochemicaland/orgeneticanalysisofbloodand/ortumortissuesamplesandintracranialactivity(ORR,DCR,DORandPFS)perResponse AssessmentinNeuro-OncologyBrainMetastasescriteria.Plasma ctDNAsampleswerecollectedonday1ofcycle1,day8ofcycle1,day1 ofcycle4,at diseaseprogressionandat EOTindose-escalationcohorts. Forpatientswhounderwentdoseescalation,additionalplasmactDNA sampleswerecollectedpredoseonthedayofdoseescalationandon day 21lafterdose escalation.AdditionalctDNA samples were alsocollectedonday1ofcycles8and16indose-expansioncohorts.Acquired genealterations and KRASstatus were assessedatEOT.AcquiredalterationsweredefinedasthosedetectedatEOTbutnotatbaseline.Patients wereeligibleforanalysisofacquired alterationsiftheyhaddisease progressionand hadasamplecollectedforctDNAanalysisatthetime ofprogressive disease or EOT.

Brain scans(MRI preferred)wereperformed atbaselinefor patientswithknownorsuspectedbrainmetastases.Forpatientswith brainmetastasesatbaseline,post-baselinebrainscanswereperformed every6weeksuntilweek25,thenevery12weeksthereafter.Forpatients withoutbrainmetastases atbaseline,brainscans wereperformed as clinicallyindicated.

Biomarkeranalysis

Next-generationsequencing(NGS) of plasma-derived ctDNA was performed using Guardant360 CDx, which is a Clinical Laboratory ImprovementAmendments-certified,CollegeofAmericanPathologistsaccredited,NewYorkStateDepartmentof Health-andUSFood and DrugAdministration-approvedclinical NGSlibrarydevelopedby Guardant Health.Plasma isolation,circulatingfree DNA(cfDNA) extraction,libraryconstructionandsequencing,andqualitycontrolassessment were performed aspreviously described35.36.Variants wereidentifiedwithavalidatedcustombiomarkerspipelinethatuses molecular barcoding and double-stranded consensus sequencing to achieve an analytical and clinical specificity >99.5% and sensitivity {>}85.0% Singlenucleotidevariantsandsmallinsertionsordeletionsin 74 genes,copy numberamplifications in18genes,andgenerearrangementsorfusionsinsixgeneswereassessed.Inaddition,theNGSpanel hasbeen validated for microsite instability assessment37.Germline alterations wereexcludedaspreviously described35.36

For sites in mainland China only, NGS was conducted by Burning Rock using an OncoCompass Target panel.DNAisolation and targeted sequencingwereperformed atBurning RockBiotech,acommercial clinical laboratory accredited bytheCollege of AmericanPathologists andcertified bythe Clinical LaboratoryImprovementAmendments,accordingto optimized protocols as described previously38.39. cfDNAfromplasmasamplewasextractedusingQIAampCirculating Nucleic Acid Kit (Qiagen) for manual extraction.Quantification of cfDNA was performed using the Qubit 2.0 Fluorometer with the dsDNA HS assaykit(Life Technologies).A minimum of10 ngofcfDNA was required for NGS library construction.Library preparation and enrichment processes were performed using OncoCompass Target Cancer MutationProfiling Liquid Kit(Burning Rock Biotech)with minormodifications.Briefly,forplasma cfDNA, 80ng wasused forlibrarypreparationwhen availableabove80ng;otherwise all extracted DNA wasused when theyieldwas10-80ng.DNA was subjected to end repair, phosphorylation and adapter ligation,followed bypolymerase chain reaction amplification.Thepurified pre-enrichment library was hybridized to an OncoCompass Target (Burning Rock) panel covering 168 human cancer-related genes, followed byhybridselectionwithstreptavidin-coatedmagnetic beads,andpolymerasechainreactionamplification.Abioanalyzer high-sensitivity DNA assay was subsequently performed to assess the quality and size ofthe library.Indexed samples were sequenced on the Novaseq60o0 platform (lllumina) with pair-end reads. Sequencing data were mapped to the human genome (hg19) by using Burrows-Wheeler Aligner (v.0.7.10). Local alignment optimization, variantcallingand annotationwereperformedwiththeGenome Analysis Toolkit(v.3.2.2),Vardict(v.1.5.1)and VarScan(v.2.4.3). Variants were filtered by using the VarScan (v.2.4.3) filter pipeline and selectedexons and intronsof168geneswerecaptured.Variants werefurtherfiltered accordingtothe populationfrequencyfrom authoritative databasesincluding ExomeAggregationConsortium, 1000 GenomesProject,dbSNP and ESP6500SI-V2.Remaining variants were annotated with the ANNOVAR (v.20160201) and SnpEff (v.4.3g) software.Structural rearrangement was analyzed using an in-housealgorithm marksV(Burning Rock Biotech)\*0.Briefly,markSV is analgorithmforanalyzingstructuralvariation,whichintegrates split-readingandpaired-endanalysiswhichissuitablefordetecting deletions,tandem duplicationevents,inversions and translocations. Thecopy numbervariation was detected by in-house analysis scripts basedondepthofcoverage dataofcaptureintervals.Coverage data wasfirstly correctedofsequencingbiasresultingfromGCcontent and probe design.The average coverage of all capture regions was used tonormalizethecoverage of different samples to comparable scales.Copynumberwascalculatedbasedontheratiobetweendepth of coverageintumorsamples and average coverage of an adequate numberof samples withoutcopy numbervariation asreferences as toeachcaptureinterval39.Genecopynumbervariationwasreported ifthe coverage data ofthe generegion wasquantitatively and statistically significantly differentfrom itsreference control.

Statistical analysis

Thesamplesizeforthetrialwasnotbasedonstatisticalhypothesis. The sample sizeswerebased onpractical considerationsconsistent withoncologystudies,withtheobjectivetoestimatetheMTDandto evaluateinitialsafetyandtolerability.Safetywasassessedinallpatients whoreceivedatleastonedoseofD3S-o01(fullanalysisset).Pharmacokineticswere assessedinallpatients whoreceived any amountofstudy treatmentandhad atleastonepharmacokineticmeasurementafter dosing.Responseendpointswereevaluatedinallpatientswhoreceived anyamountofstudytreatment,hadmeasurablediseaseatbaselineand at least onepost-baselineimaging result(response-evaluableset).All outcomeswereassessedusingdescriptivestatistics.Clopper-Pearson exact 95% CIswerecalculatedforORRandDCR.ForlandmarkDORand PFS, 95% ClswereestimatedbyKaplan-Meiermethodology.Datawere collected using iMedidata RAVE EDC.Analyses were performed using commercially availablesoftware(SASv.9.4).Figures were produced usingRv.4.2(http://www.R-project.org/)and GraphPadPrismv.10.2.2 forWindows(www.graphpad.com).This studyisregistered with ClinicalTrials.gov,NCT05410145.

Reporting summary

FurtherinformationonresearchdesignisavailableintheNature PortfolioReportingSummarylinkedtothis article.

Data availability

Deidentifiedparticipantdatathatunderlietheresultsreportedin thisarticle,theprotocolandstatisticalanalysisplanmaybeavailable uponreasonablerequesttothecorrespondingauthorfromqualified researchersfollowingcompletionofthestudywhenalldose-expansion cohorts arecompleted and theclinicalstudyreportshavebeenfinalized.Dataareonlyavailableuponrequesttoprotecttheprivacyofthe company and clinicaltrialparticipants.Qualified researchers mayemail Publication@d3bio.comtorequestdata.Aresponsewillbeprovided within2monthsoftherequest dependingontheextentoftherequest. Databasesused inthis studywere Exome Aggregation Consortium (http://exac.broadinstitute.org),1000GenomesProject(https://www. internationalgenome.org/),dbsNP(https://www.ncbi.nlm.nih.gov/ snp/)and ESP6500SI-V2(https://genome.ucsc.edu/).

References

35.Odegaard,J.l.et al.Validation ofa plasma-based comprehensive cancergenotyping assayutilizingorthogonaltissue-and plasma-based methodologies.Clin.CancerRes.24,3539-3549 (2018).
36.Lanman,R.B.et al.Analytical and clinical validation of a digital sequencingpanelforquantitative,highlyaccurateevaluation of cell-free circulating tumor DNA.PLoS ONE10,e0140712 (2015).
37.Willis,J.etal.Validation ofmicrosatellite instabilitydetection using a comprehensive plasma-based genotyping panel. Clin. CancerRes.25,7035-7045(2019).
38.Mao, X.et al. Capture-based targeted ultradeep sequencing inpairedtissueandplasma samplesdemonstratesdifferential subclonal ctDNA-releasing capability in advanced lung cancer. J. Thorac. Oncol. 12, 663-672 (2017).
39.Li, Y.S.et al.Unique genetic profiles from cerebrospinal fluid cell-freeDNAinleptomeningealmetastasesofEGFR-mutant non-small-cell lung cancer: a new medium of liquid biopsy. Ann.Oncol.29,945-952(2018).
40.Tao,Z.et al.Profiling receptor tyrosine kinase fusions in Chinese breast cancers.Front.Oncol.11, 741142(2021).

Acknowledgements

FundingforthisresearchwasprovidedbyD3Bio.Wethankthe patientswhoparticipated inthistrial,aswellastheirfamilies and caregivers,and thestaffand investigators atallthetrialsites.Medical writing assistance was provided by M.Sweetlove and S.Duggan of ApotheCom,andwasfundedbyD3Bio.

Author contributions

B.C.C., S.L.,T.M., Z.J.C., C.C.and J.W.were responsible for the study cOnception or design.B.C.C., S.L.,M.A.L.,Z.S.,J.J.P., S.M.L.,Z.L., J.Zhao, G.R.,Y.Z., Jun Zhang, A.L.,H.H.L., C.C., J.W.,Y.S.,Z.F., Q.C., H.W.,Jing Zhang,Z.J.C.,M.L.J.and T.M.conducted the study and were involved in acquisition, analysis or interpretation of data.J.W. performedstatisticalanalysis andY.S.verified thebiomarkerdata. B.C.C.,S.L., M.A.L.,Z.S.,J.J.P., S.M.L., Z.L., J.Zhao, G.R., Y.Z., Jun Zhang, A.L.,H.H.L., C.C.,J.W.,Y.S.,Z.F., Q.C., H.W.,Jing Zhang,Z.J.C., M.L.J. and T.M.draftedtheworkorrevieweditcriticallyforimportantintellectual COntent.B.C.C., S.L., M.A.L.,Z.S., J.J.P., S.M.L.,Z.L., J.Zhao, G.R., Y.Z., Jun Zhang, A.L., H.H.L.,C.C., J.W.,Y.S., Z.F., Q.C., H.W., Jing Zhang, Z.J.C.,M.L.J.and T.M.approved thefinalversionforpublication and agreetobeaccountableforall aspectsof theworkinensuringthat questionsrelatedtotheaccuracyorintegrityofanypartofthework are appropriatelyinvestigated andresolved.

Competing interests

B.C.C.reportsresearchfunding from Gllnovation,AstraZeneca, ChampionsOncology,CJBioscience,Cyrus,Janssen,MerckSharp& Dohme,Dong-A ST,Yuhan,ImmuneOncia,Therapex,JINTSbio and VerticalBioAG;royaltiesfromChampionsOncology,Crown Bioscience,Imagen and PearlRiver BioGmbH;consulting fees from BeiGene,ovartis,AstraZeneca,BoehringerIngelheim,Roche,Bristol Myers Squibb,CJ,Cyrus Therapeutics,Ono,Yuhan,Pfizer,Eli Lilly, Janssen,Takeda,MerckSharp&Dohme,Gilead,Amgen,Daiichi Sankyo,Regeneron,anofi,AnHeartTherapeutics,eagen,Harpon Therapeutics,GlaxoSmithkline andArriVent;honorariaasaninvited speakerfromASCO,AstraZeneca,Guardant,Roche,ESMO,IASLC, Korean Cancer Association, Korean Society of Medical Oncology, Korean Society of Thyroid-Head and Neck Surgery,Korean Cancer StudyGroup,Novartis,MerckSharp&Dohme,theChineseThoracic OncologySociety,Pfizerand ZaiLab;participationonscientific advisory boards forKANAPH Therapeutic Inc.,BridgebioTherapeutics, CyrusTherapeutics,GuardantHealth,JINTSBioand TherapexCo.Ltd; stockownershipwithTheraCanVacInc.,GencurixInc.,Bridgebio Therapeutics,KANPHTherapeuticInc.,CyrusTherapeutics,Interpark Bio Convergence Corp.andJ INTS Bio;and other financial or nonfinancialinterestswithYonsei UniversityHealthSystem (employment),DAAN Biotherapeutics(founder)and JINTS BIO (memberoftheboard ofdirectors).S.L.reportsresearchfunding from AstraZeneca,Hutchison,BristolMyers Squibb,HengruiTherapeutics, BeiGene,Roche and Hansoh;consulting fees from AstraZeneca, Hutchison,Simcere,Zailab and Yuhan Corporation;honoraria from AstraZeneca,Roche,Hansoh and Hengrui Therapeutics;participation onadvisoryboardsforAstraZeneca,YuhanCorporation,InventisBio， MerckSharp&Dohme,SimcereZaiming Pharmaceutical,Shanghai FosunPharmaceuticalandPhaseTherapeutics;andaleadershiprole with Innovent Biologics. Z.L.reports honoraria from AstraZeneca, Roche,Hansoh andPfizer.G.R.reportsresearchfundingfromBristol MyersSquibb,Roche/Genentech,AstraZeneca,Merck,Takeda， BeiGene,Pfizer,CBTPharmaceuticals,CorvusPharmaceuticals, Novotech,Shanghai FosunPharmaceutical Development,Shanghai Henlius Biotech,Five Prime Therapeutics,Suzhou Alphamab, BoehringerIngelheim,Adagene Inc.,Bio-TheraSolutions, ChemoCentryx,CuronBiopharmaceutical,D3Bio,InventisBio, SenzOncology,GenfleetTherapeutics,GeneQuantumHealthcare, Keythera Pharmaceuticals,LaNova Australia,MedicennaTherapeutics,

Australia,RemeGen,Seagen,Surface Oncology,Eucure Biopharma, Janssen Oncology,ImmuGen, Imugene, Therapim,Zentalis and Agenus.Jun Zhang reports participation on advisory boards for Astellas and Shanghai HengRui. H.H.L. reports research funding from MSD, MundiPharma and Novartis; consulting fees from Boehringer Ingelheim,Celgene,EliLilly,Illumina,Novartis,MerckSerono,Takeda and GeorgeClinical;honorariafromAbbVie,Bayer,Eisai,EliLilly, Guardant Health and Novartis; and support for meeting attendance and/or travel fromBayer,BoehringerIngelheim,Merck Sharp& Dohme, Novartis and Pfizer. C.C., J.W., Y.S., Z.F., Q.C., H.W., Jing Zhang and Z.J.C.reportemployment and stockorstockoptions withD3Bio. M.L.J.reports funding to the institution from D3 Bio for the current work;grants tothe institutionfromAbbVie,Adaptimmune,Amgen, Arcus Biosciences,ArrayBioPharma,ArriVent BioPharma,Artios Pharma,AstraZeneca,Bayer,BeiGene,BerGenBio,BioAtla,Black Diamond,BoehringerIngelheim,Bristol MyersSquibb,Calithera Biosciences,CarismaTherapeutics,Cityof HopeNationalMedical Center,ConjuproBiotherapeutics,CorvusPharmaceuticals,Curis, CytomX,DaiichiSankyo,DracenPharmaceuticals,Lilly,Elicio Therapeutics,EMD Serono,EQRx,Erasca,Exelixis,FateTherapeutics, Genentech/Roche,Genmab,Genocea Biosciences,GlaxoSmithKline, Gritstonencology,Harpoon,Helsinn ealthcareSA,Hengrui Therapeutics,HutchinsonMediPharma,IDEAYABiosciences,IGM Biosciences,Immuneering Corporation,Immunitas Therapeutics, Immunocore,Impact Therapeutics, Incyte, Janssen, Kartos Therapeutics,LockBodyTherapeutics,LoxoOncology,Memorial SloanKettering,Merck,Merus,MiratiTherapeutics,Mythic Therapeutics,Neolmmune Tech,Neovia Oncology,NextPoint Therapeutics,Novartis,NumabTherapeutics,Nuvalent, OncoC4, PalleonPharmaceuticals,Pfizer,PMVPharmaceuticals,Rain Therapeutics,RasCalTherapeutics,Regeneron Pharmaceuticals,Relay Therapeutics,Revolution Medicines,Ribon Therapeutics,Rubius Therapeutics,Sanofi,SevenandEightBiopharmaceuticals/Birdie Pharmaceuticals,Shattuck Laboratories,SiliconTherapeutics, Systimmune, Taiho Oncology,Takeda Pharmaceuticals, TCR2 Therapeutics,Tempest Therapeutics,TheRas, Tizona Therapeutics, TMUNITYTherapeutics,Turning Point Therapeutics,Vividion,Vyriad andY-mAbsTherapeutics;consultingfeesfromAbbVie,Alentis Therapeutics,Amgen,Arcus Biosciences,AstraZeneca,Biohaven Pharmaceuticals,BoehringerIngelheim,BristolMyers Squibb,D3Bio Limited,Daichi Sankyo,Fate Therapeutics,Genentech/Roche,Gilead Sciences,GlaxoSmithkline,GritstoneOncology,HookipaBiotech, Immunocore,JanssenPharmaceuticals,Lilly,Merck,Mirati Therapeutics, ModeX Therapeutics,Normunity, Novartis, Novocure, Pfizer, Regneron Pharmaceuticals, Revolution Medicines, Sanofi-Aventis,SeaGen,Synthekine,TakedaPharmaceuticals andZai Laboratory.T.M.reports research funding from AstraZeneca,Bristol Myers Squibb,G1Therapeutics,Merck Sharp &Dohme,Merck Serono, Novartis,Pfizer,Roche,SFJ,Takeda andXCovery;consultingfeesfrom AbbVie,ACEAPharma,Adagen,AlentisTherapeutics,Alpha Biopharma,Amgen,Amoy Diagnostics Co,AnHeart Therapeutics, AVEO Pharmaceuticals,Bayer Healthcare Pharmaceuticals,BeiGene, BerGenBio ASA,BerryOncology,BoehringerIngelheim,Blueprint MedicinesCorporation,BridgeBio,BristolMyersSquibb,Bowtieif Insurance Company Limited,Bridge Biotherapeutics Inc., Covidien LP, C4 Therapeutics Inc., Cirina,CStone Pharmaceuticals,Curio Science, D3 Bio,Da Volterra,Daichi Sankyo,Eisai,Elevation Oncology, Erasca,F.Hoffmann-LaRoche/Genentech,FishawackFacilitate, G1 Therapeutics,geneDecode, Gilead Sciences,GLG's Healthcare, Gritstone Oncology, Guardant Health, HengRui Therapeutics, HiberCell, HutchMed, Ignyta,Illumina, magene Al, Incyt Corporation,Inivata,InxMed,IQVIA,Janssen,LakeshoreBiotech,Lilly, Lunit USA, Loxo Oncology, Lucence Health, Medscape/WebMD,

Medtronic,MerckSerono,MerckSharp&Dohme,MiratiTherapeutics, MiRXES,MoreHealth,Novartis,Novocure GmbH,NingboNewBay Technology Development,OmegaTherapeutics,OrigiMed,OSE Immunotherapeutics,PeerVoice,PhanesTherapeutics,Pfizer,PrIME Oncology, Prenetics Global, Puma Biotechnology, Qiming Development(HK),Regen Medtech Holdings,Regeneron Pharmaceuticals,RochePharmaceuticals/Diagnostics/Foundation One,Sanofi-Aventis,Schrodinger,SeagenInternationalGmbH,SFJ Pharmaceutical,SimcereofAmerica,SynergyResearch,Summit Therapeutics,TakedaPharmaceuticals HK,Tigermed,Vertex Pharmaceuticals,Virtus Medical Group,XencorandYuhan Corporation;honoraria from ACEA Pharma,Alpha Biopharma,Amgen, AmoyDiagnostics,AstraZeneca,BeiGene,BoehringerIngelheim, BristolMyersSquibb,DaiichiSankyo,DazGroup,FishawackFacilitate, InMedMedicalCommunication,JanssenPharmaceuticaNV,Jiahui Holdings,LiangYiHui Healthcare,Lilly,LucenceHealth,MD Health Brazil,Medscape,MerckPharmaceuticalsHK,MerckSharp&Dohme, MiRXES,Novartis,OrigiMed,P.Permanyer SL,PeerVoice,Physicians EducationResource,Pfizer,PrlMEOncology,ResearchtoPractice, RochePharmaceuticals/Diagnostics/Foundation One,Sanofi-Aventis, Shanghai BeBirds Translation & Consulting,Taiho Pharmaceutical, Takeda Oncology,andTouchIndependent Medical Education;support formeeting attendance and/or travel from Novartis,Roche,Pfizer, AstraZeneca,DaichiSankyo,BoehringerIngelheim,MiRXES,Bristol MyersSquibb,Merck Sharp&Dohme,AbbVie,ZaiLab and Liangyihui; participationin advisoryboardsforAbbVie,ACEAPharma,Amgen, AstraZeneca,AlentisTherapeuticsAG,BerGenBioASA,Berry Oncology,lueprint MedicinesCorporation,BoehringerIngelheim, BowtieLife Insurance,Bristol Myers Squibb,C4Therapeutics,Covidien LP, CStone Pharmaceuticals,Curio Science,D3 Bio,Daichi Sankyo, Eisai,Erasca,FishawackFacilitate,G1Therapeutics,GileadSciences, Gritstone Oncology, Guardant Health, geneDecode (uncompensated), Hengrui Therapeutics, HutchMed, Ignyta, Incyte Corporation,Imagene Al InivataQnenLartchillxn Lunit,MerckSerono,MerckSharp&Dohme,MiratiTherapeutics, MiRXES Group,Novartis,OrigiMed,PhanesTherapeutics,Pfizer, Prenetics Global,PumaBiotechnology,Roche/Genentech,Regeneron Pharmaceuticals,Sanofi-AventisR&D,SFJPharmaceutical,Simcereof America,SimcereZaiming,Takeda,VertexPharmaceuticals,Virtus Medical Group,Xencor and Yuhan Corporation;leadershiprole for AstraZeneca,HutchMed,Aurora,Epoch Biosciences and Insighta; stock or stock options with AstraZeneca,AuroraTele-Oncology, Biolidics,HutchMed,Prenetics Global,D3Bio,Lunit,BowtieLife Insurance,LakeshoreBiotech,Loxo Oncology,Virtus Medical Group, Yinson Capital,Phanes Therapeutics,Insighta and Alentis Therapeutics AG.The other authors declare no competing interests.

Additionalinformation

Extended data is available forthis paper at https://doi.org/10.1038/s41591-025-03688-6.

Supplementary information The online version contains supplementary materialavailableathttps://doi.org/10.1038/s41591-025-03688-6.

Correspondence andrequestsformaterialsshouldbeaddressedto Byoung Chul Cho.

Peer reviewinformation Nature Medicine thanks Benjamin Besse, DavidHong and theother,anonymous,reviewer(s)fortheir contributiontothepeerreviewof thiswork.PrimaryHandlingEditor: Saheli Sadanand, in collaboration with the Nature Medicine team.

Reprintsandpermissionsinformationisavailableat www.nature.com/reprints.

ExtendedDataFig.1/D3S-001pharmacokineticsatsteadystateinpart 1(phase ladose escalation).(A) Geometric mean steady-state plasma concentrations;errorbarsrepresentgeometricstandarderror.(B)Individual

steady-stateplasma \mathsf{A U C}_{\mathsf{t a u,t o t a l}}, Thesamplesizeforeachdosegroupis: 50mg (n=2) ;100 mg (n=7) ;200mg (n=4) ;400mg (n=9) 600mg(n=8) 900{mg} (n=8) .AUC,areaundertheconcentrationtimecurve;PK,pharmacokinetics.

ExtendedDataFig.2|Magneticresonanceimagingscansofbrainand computedtomographyscansoflivermetastaticlesionsfollowingtreatment withD3s-oo1.Magneticresonanceimagingscansofthebrainandcomputed tomographyscansoftheliverareshownforapatientwithNsCLCwhohad priortreatmentwithplatinum doubletchemotherapyandimmunotherapy (atezolizumab).Thepatienthadpreviouslytreated/stablebrainmetastatic lesionat baseline(received Gamma Knife treatment1.5yearsbefore starting

studytreatmentwithD3S-001duringwhichtimenopostradiationshrinkage was observed).FollowingD3s-001,the patient hadconfirmed partialresponse, andshrinkage ofthebrainmetastaticlesionwas observedfollowingthefirst dose \mathbf{(50mg)} withfurthershrinkageobservedinparallelwithshrinkageofliver metastaticlesionathigherdosesduringintra-patient doseescalation.Treatment with D3S-001isongoing( >22 months).

ExtendedDataTable5|PriorKRAS G12Ci treatment inpart2(phase1b dose-expansioncohort)

nature portfolio

Reporting Summary

Naturerltitfttliisitsister inreportinrfurthnftinnaturrflplcieeurdiriallcinthitorilPolihklis

Statistics

Ourwebcollection onstatisticsforbiologistscontainsarticles onmanyof thepoints above.

Software and code

Policy information about availability of computer code

Data collection iMedidata RAVE EDC.

Data analysis

SAS v 9.4, SAS Institute, Cary, NY, USA), R v4.2, and GraphPad Prism v10.2.2 for Windows.

asr eesef

Data

Policyinformationaboutavailability ofdata

Allmanuriptitabitastet hldrdethllntheli -Accession codes, unique identifiers, or web links for publicly available datasets -A description of anyrestrictions on data availability -For clinical datasets orthird partydata,please ensure that thestatement adheresto ourpolicy

Deidesa theorennqualrchsflwntthuhlxpareleathls reprthtu maildqtwfthqdintqba

Research involving human participants, their data, or biological material

Note that fullinformation onthe approval of the study protocolmust alsobe provided inthe manuscript.

Field-specific reporting

easettitdtior

Life sciences

Behavioural&socialsciences

Ecological,evolutionary&environmentalsciences

Forareferencecopyofthedocumentwithallsectionsseeaturecom/documents/nr-reporting-summary-fladf

Life sciences study design

All studies must disclose on these points even when the disclosure is negative.

Reporting for specific materials, systems and methods

Werq stdlife

Materials & experimental systems Methods

n/a Involved in the study ± ChIP-seq ± Flow cytometry ± MRI-based neuroimaging

n/a Involved in the study
α Antibodies
α Eukaryotic cellines
α Palaeontology and archaeology
α Animals and other organisms
口 α Clinical data
α Dual use research of concern
α Plants

Clinical data

Policyinformation aboutclinical studies

Almaurilthiafichdekiswis

Clinical trial registration

Study protocol

Data collection

Outcomes

Plants

Seed stocks

Novel plant genotypes