際際滷

際際滷Share a Scribd company logo

Commercialisation of Cellar Therapies for Cancer (final)

Download as DOCX, PDF
awalker112526
awalker112526

This document discusses the challenges and opportunities for commercializing cellular therapies like CAR-T cell therapies. It notes that while CAR-T therapies have shown spectacular results for some blood cancers, significant hurdles remain for applying them to solid tumors which represent the vast majority of cancers. Manufacturing CAR-T cells is a complex, individualized and expensive process that faces challenges in scaling. For CAR-T therapies to be commercially viable, they will need to demonstrate durable and high rates of remission for a wide range of cancers at a cost that healthcare systems are willing to pay, which may be in the range of $150,000-$300,000 per patient. New technologies that improve the efficacy, safety and scalability of cellular

1 of 7
Download to read offline
Commercialisation of Cellular Therapies Anthony Walker, PhD and Robert Johnson Abstract Successful commercialisationof acell therapyrequiresmore thanprovingsafetyandefficacytothe regulators. The inherentcomplexityof cellularproductsdelivers particularmanufacturing,logistical and reimbursementhurdlesthatthreatencommercial viabilityforanytherapywithalessthan spectacularclinical profile thattrulychangesthe standardof care. Thisisparticularly acute for autologouscell therapieswhere patientsreceive bespoketreatmentsmanufacturedfromasample of theirowncellsandwhere economiesof scale,whichplayanimportantrole incontainingthe productioncostsfor small molecule andantibodytherapeutics,are highlylimited.Nevertheless,the promise of game-changingefficacy,asexemplifiedbyveryhighlevelsof complete responsesin refractoryhaematological malignancies,hasattractedcapital investmentsonavastscale,and the attendantpace of technologydevelopmentprovidespromisingindicatorsforfuture clinical and commercial success. Keywords:cellulartherapy,CAR-T,solidtumours,immunotherapy,oncology The CAR-T Landscape Clinical trialshave revealedspectacularefficacyinrefractoryleukaemiapatientsusinganti-CD19 chimericantigenreceptor(CAR) adoptiveT-cell therapy,witha92% complete remissionrate in39 patientswith Acute LymphocyticLeukaemia(ALL).These are transformative resultswhichhelptip the balance inthe waragainst cancer:if the complete responsesare durable,the technologyraises the previouslyunthinkablepossibilitythatlate stage cancercan be cured. These andotherdata have ignitedinvestorinterestinthe biotechsectorandthisfieldinparticular. Start-upcompanieshave raisedunprecedentedamountsof private capital,subsequentlytapped equallyimpressive sumsinIPOsand have gone onto achieve eye-wateringlyhighvaluationsonthe publicmarkets.Suchisthe appetite thatimportantinvestmentcriteria, forexample robustandclear IP have beensomewhatforgiven,althoughthishasbeenshiftingasthe sectorrapidlymatures. Big pharmahas alsodivedintothe field;inAugust2012 Novartisenteredintoanalliance withthe Universityof Pennsylvania(UPenn) onthe backof nascentclinical datawithU PennsCD19 construct.Novartisagreedtofunda CenterforAdvancedCellularTherapies(CACT) onthe U Penn campusin Philadelphia,contributing$20m.At the time of the deal,thiswasan extraordinary venture forNovartiswhichmaintainedastrongfocuson small moleculeandmonoclonal targeted therapies;throwingitsweightbehindanautologouscelltherapywasaradical departure forits typicallyconservative approachtodeal-making.The companyislookingtoaggressivelymaintainits leadershipinthe fieldthroughbothin-houseresearchandcontinueddeal-making. In March 2013 Celgene andBluebirdBioannouncedaglobal strategiccollaborationtoadvance gene therapyinoncology,withafocus on CAR-Technology.Financialtermsof the agreementinclude an upfrontpaymentandup to $225 millionperproductinpotential optionfeesandclinicaland regulatorymilestones. In June 2014 PfizerandCellectisenteredintoaglobal strategiccancerimmunotherapycollaboration to developCAR-Ttechnologies.PfizergainedaccesstoCellectisallogeneicapproachandPfizeris
able to select15 targets.Cellectisreceivedanupfrontpaymentof $80m and milestonepaymentsof up to $185m. In January2015, AmgenandKite announcedastrategiccancer immunotherapycollaboration, combiningAmgensoncologytargetsandKitesCAR-Tplatformtodevelopnewtherapeutics.Kite received$80m upfrontfromAmgenandiseligibleforupto $525m in milestonesperprogram. Celgene redoubleditscommitmenttothe fieldinJune 2015, announcinga ten-yearcollaboration withJuno(albeitaftersignificantlynarrowingitscollaborationwithBluebird).Celgenegained optionstocommercialize Junosprogramsoutside NorthAmericaandmade aninvestmentof $1bn. However,the technologyfacesbigquestionstolive uptostratosphericexpectations,withatleast twomajor issuesfacingthe field.First,canthese technologiesbe commercialisedatreasonablecost (CAR-Ttechnologiesuse autologouscells,amajorlogistical challenge),andsecond,canthe technologybe appliedtocancersbeyondleukaemia.A technologythatdeliversequallyimpressive resultsina range of solid tumours wouldsurelymarkamilestoneinthe historyof medicine. Addressing Solid Tumours Althoughsuccessfultreatmentof advancedhaematological malignancieswouldrepresentamajor clinical win,commercial success relies onconqueringsolidtumourswhichrepresentover90% of the oncologysector.There isa dearthof clinical data;the mostimpressive todate include: A 60% response rate intensynovial sarcomapatientstreatedwithNY-ESO-1SpecificT-Cells producedbyAdaptimmune,withadurabilityof responserangingfromtwotonine months [1]; The U Penngrouptreatedpatientswithpancreaticductal adenocarcinoma(PDAC) with CAR-T cellsrecognisingmesothelin,anantigenoverexpressedonPDACcells.Twoof six patients experiencedstable disease byRECIST1.1 withdisease control off-therapyseeninone patientfor > 4 months.In one patient,abnormal 18FDG avidityseeninlivermetastasesatbaselinewasno longerdetectedat1 monthaftertherapy [2]. Whilstprovidingencouragingpreliminarysignsof efficacy,these resultsfall well shortof the possibly un reasonable game-changing benchmarksthatmanyinvestorsandsome BigPharmahave come to expectfromadoptive cellularimmunotherapy. Major technical challengesremain, notleasthomingof sufficientengineeredTcellstothe tumour site(s) toprovide anadequate efficacy:toxicity ratioforefficacy.Beyondthis,the microenvironment of a solidtumourmayprove to be far more hostile to CAR-Tcellsthancirculatingtumours,and inductionof anergy maybecome anefficacy-limitingfactorregardlessof CAR-Tcell persistence whichinitself maybe an issue.Finally,the emergence of CAR-Tresistantdisease throughantigen escape lossmechanismsalreadyseeninstudiesof cancervaccines [3] may limitthe durabilityof any responsesto CAR-Ttherapyand,as discussedbelow,thiscouldseverelyrestrictreimbursementfor thismodality. Manufacturing & Logistical Hurdles Most CAR-Tand TCR-engineeredTcells are currentlymade bya cumbersome andbespoke process involving: LeukapheresistoextractT cellsfroma cancer patientwhoisconnectedbytwointravenous tubesto an apheresismachineforseveral hours.Thisisnotcomfortable forthe patient,incursa
substantial cost,andultimately,large-scaleadoptionof autologous CAR-Ttherapymaybecome rate limitedbyavailabilityof apheresiscapacity; Activationandtransductionof T cells.The original CAR-Tprocessesusedanti-CD3anti-CD28 DynabeadsforT cell activation;Novartismade anearlymove tosecure exclusive commercial access to thisreagentforCAR-Tproductiontoreinforce itsproprietarypositioninthe field.Since then,alternativestoDynabeadshave beensuccessfullyimplementedandthisnolonger representsacommercial hurdle.Transductionisusuallybyretroviral orlentiviral vectors, althoughnon-viral systemsare alsoused; Expansionof transducedTcellsoveran approximatelytwo-weekperiodinacytokine (typically IL-2) supplementedtissueculture medium; Washingand concentratingthe Tcellspriorto administration.For CAR-Tproducts made at central facilitiesandtransportedtoremote treatmentcentres,cryopreservationprotocolshave beendeveloped; QC release assaysare conductedforeachbatch of CAR-Tproduct. The entire processhasto be conductedunderenvironmentallycontrolledGMP compliantconditions whichare expensivetomaintainandrun.Aseach CAR-Tproduct ismade fromstartingmaterials(T cells) fromthe patienttobe treated,there are nosubstantial economiesof scale:increasing manufacturingcapacityrequiresthe additionof extraproductionpodsorboothsinparallel.The earliestproductionprocesseswere lab-based,fullymanual andinvolvedopentissueculture manipulationsinbiological safetycabinetsand/orisolators:althoughworkable forveryearlystage clinical research,thisapproachisclearlynotappropriate forcommercial purposes.Anindustrial grade manufacturingprocessneedstobe conductedinsealedsystems(zeroopenmanipulations) and shouldbe automatedasfar as possible,primarilytoreduce the humanvariabilityfroma complex processbutalsotoreduce costs. A fullysealedsystemcouldarguablybe situatedinaGrade D cleanroom(EU cGMP classifications),althoughindividual regulatoryagencieswill scrutinisethisin great detail andmayrequire Grade C conditionswhichincreasescomplexityandcost.Atpresent, the onlycommerciallyavailablesystemthatmeetsthe criteriaof a closedsystemautomated manufacturingdevice isthe Miltenyi CliniMACSProdigy [4]. Eventhen, noteveryprocesscanbe implementedonthatsystem(forexample, itappearsbettersuited tolentiviral transductionthan retroviral,andthisdependsonthe detailsof each specificprocesswhichwill notbe discussedhere). There are twoschoolsof thoughtregarding the locationof CAR-Tmanufacturing.The BigPharma industrial mind-setfavourslarge,centralisedfacilitiescapableof processingTcellsfrommany thousandsof patientsperyear.The alternative isadecentralisedmodel withmanufacturingco- locatedwiththe specialisttreatmentcentres(e.g.universityhospitals,majorcancertreatment centres).Theoretical modelsof the scalingof anautologousproductionmodulecanshow thatthere are optimumsize thresholdsforaproductionmodule atdifferentdose numbersperyear. Economiesof scale are non-linear:i.e. there are scenarioswheretwosmallproductionmodulesin parallel,eachcapable of producingsay500 dosesperyear,are more efficientthanamedium module scalesfor1,000 dosesperyear.These considerationsclearlyplayalarge role inthe decision betweenthe twomodels(andindeed the degreeof decentralisation;i.e.whatwouldthe optimum numberof sitesbe inthe USA; 25, 50, 100?). Top level pros andconsof eachmodel include are listedinTable 1.
Pricing, Reimbursement & Market Access Whichevermanufacturingmodel ischosen,manufacturingcostsforautologous CAR-Tcell therapies are likelytobe inthe range of $25-35,000 perpatient,even aftermaximumprocessefficiencieshave beenexploited.Note thatthisisjustthe cost of manufacturingthe therapy,andinfact a course of treatmentincurssubstantial additional costsforpatientpreparation(mostcurrentprotocolscall for lymphodepletionpriorto CAR-Tadministration,anexpensive andtime consumingin-patient procedure), CAR-Tadministrationandfollow upcare costs(notablythe resourcesrequiredfor monitoringandtreatingthe severe sideeffectse.g.cytokine stormoftenassociatedwithCAR-T therapy). Total treatmentcostsmaytherefore be ashighas those for autologous bone marrow transplantation (currentlyca.$360,000 in the USA) plusthe price of the CAR-T therapy. In the immediate aftermathof the Novartisdeal with UPenn,financialanalystswereestimatingthat CAR-Ttherapieswouldbe pricedat$250,000 per patientandpossiblyhigher. AsnoCAR-Tproducts have beenlaunchedontothe marketyet, future pricingremainsspeculative,butmany analystsare now basingforecasts onprice ranges between$150,000 and$300,000 per patient.Eventhe lower endof thisrange isabove current annualisedmedicationcostsfornewly-launchedcancerdrugs. At firstglance,this appearstoflyin the face of the growingdebate aboutthe JustumPretium,orJust Price,of cancer drugs [5], voluble clinicianresistancetoperceivedexcessive drugpricing(arguably startedinOctober2012 whenphysiciansatthe Memorial Sloan-KetteringCancerCenterannounced inthe NewYorkTimesthat theirhospital wouldntbe usingZaltrap,anew VEGF-targetedcancer treatmentformetastaticcolorectal cancer) andthe increasinguse of HealthTechnologyAssessment (HTA) to restrictthe use of highpricedtreatments. A recentstudyassessinganoriginal dataset of 58 anticancerdrugsapprovedbetween1995 and 2013 foundthatlaunchprices,adjustedforinflation and drugssurvival benefits,increasedby10%,or about $8,500, peryear[6]. Thus,in 2004 bevacizumabwas launched forpatientswithlate stage colorectal cancerata price of $50,000 per treatmentepisodeandwasassociatedwithanincremental increase inoverall survival of five months.Sevenyearslaterin2011, ipilimumabwasapprovedfortreatmentof melanoma,associated withan incremental increase inlifeexpectancyof fourmonths,andlaunchedata price of $120,000 pertreatmentepisode.Itisclearlyuntenableforthistrendtocontinue indefinitelyinanalready budget-constrainedhealthcare environment. The keyquestiontherefore is:willpayersbe preparedtofund CAR-Ttreatmentatthese levels?This iskeyto understandingthe commercial prospectsforthe technologyasawhole.The answersimply isthat it will dependonthe level of clinical efficacythatcan be achievedwiththis new modality.In the most optimistic(but probablyunrealistic) scenario,asingle CAR-Tadministrationwouldinduce longtermremission,sayfive totenyears.Usingthe UKs NICEbenchmarkof 贈30,000 ($47,000) per quality-adjustedlife year,thiswould amounttoa maximumreimbursableprice of $235-470,000. Thiswouldhave to be adjusteddownwardstoaccountforthe fact thatcancer patientsqualityof life islessthanperfect,butthatstill mayallow aprice for CAR-Ttreatmentof $118-235,000. Allowingforthe observationthatNICEsapproachisamongstthe moststringentinmainstream pharmaceutical markets,andUKpricesfor pharmaceuticalstendtobe significantlylowerthaninthe US (the mainmarket) and othercountries,itispossiblethatthe HTA-acceptedpricesforCAR-T therapywill be above thisrange andtherefore inline withcurrentanalystprojections.However,any shortfall inefficacybelowadurable five-yearremissionwouldnecessarilyreducethese calculated price ranges.Thismay perhapsbe an overlysophisticatedwayof sayingthat CAR-Ttherapieswill be paidfor onlyif theyworkinthe clinic,butthe seriousconclusionisthatthe benchmarkfor commercially-viable efficacyisgoingtobe muchhigherthanfor lesscomplex modernoncology
drugssuch as monoclonal antibodies. Unless CAR-Ttherapy, initscurrentpatient-specific autologousapproach,caninduce longtermdurable remissionsinahighproportionof treated patients,the chancesof commercial successare low. New Technology to Drive Commercial Success A numberof technologies onthe horizonappearsettogreatlyimprove the clinical andcommercial prospects foradoptive cellularimmunotherapy. Some of these approachesaimatenhancingthe efficacyof CAR-Tcells,forexample byboostingthe homingof infusedcellstotumoursites,the lack of whichisbelievedtobe amajor reasonwhyefficacyof CAR-Tcellsinsolidtumoursissignificantly belowthatincirculatingtumours.There are alsoapproachesto enhancingthe endurance of CAR-T cellsnotjustin termsof the lengthof durationinthe circulation, butalsothe retentionof anactive, non-anergised phenotype.Additionally,some researchersare assessingmethodstobuildin defencesagainsttumourmechanismsthat evade ordisable the immune system;suchanti-missile missiles mayhave alot of mileage inboosting CAR-Tefficacy. Anothersetof approachesfocuson the safetyof CAR-Ttherapies.Whilstthereissome evidence that cytokine releasesyndromeisatleastassociated,if notcorrelated withclinical efficacy,there may be methodstobluntthe severityof the reactionwithouttoomuchof an efficacyhit.Reducing the needto treatthe side effectsof CAR-Ttherapyinan intensive care unitwouldclearlybe better for patients,more practical fortreatmentcentresandmuchmore attractive to payers. Perhapsthe mostfundamental approachtoimprove cellularimmunotherapywill be totransformit froma bespoke,patientspecificprocesstoa mass-producedproduct. Thiswouldreduce leadtimes for treatment(the productwouldbe available off-the-shelf), have aninherentlylowerproduction cost (allowingforreductionsinpricingwithoutoverlysacrificingprofitmargins)andwouldbe more inline withthe model of productdistributionbythe pharmaceutical industry. The firststeptoward thisisthe use of allogeneicTcellsfromhealthydonors,forexample gene editedtoeliminate the endogenousTcell receptorandthuspreventacute rejectionbythe recipient,andwhere one donor can provide enoughcellstotreatmultiplepatients.Ultimately,we foresee thatthisapproachwill extendtoimmortalisedTcell linesthatcanbe culturedinbulkon an industrial scale,eliminatingthe needforT cell donorsaltogether.Animmortalised,allogeneicmassproduced CAR-Tcell capable of inducingdurable (five yearsormore) remissionsremainsapipe dream, butatleastis onthe radar for a numberof well-resourcedresearchgroupsandthusstandsa reasonable chance of becominga reality.
Table 1: Toplevel pros andcons of manufacturingmodels Pro Con Centralised - Economiesof scale (albeitlimited) - Extremelycomplex logistics increasesriskof production failures/errors - Higherriskof productionoutages inthe eventof plant contamination,etc. Decentralised - Enablesuse of fresh,non- cryopreservedproductwhichis likelytobe inherentlysuperior - Maximiseslocal clinicianbuy-into the therapy - Potentiallyharnessesexisting investmentsincleanroomsat majorcancer centres - Needtodemonstrate comparabilityof manufacturingat each site:complex,time consuming,costly - Validatinganyprocesschange at each site multipliesthe efforts requiredvsa centralisedsetting
References 1. Merchant, S.,Cristea,M.C., Stadtmauer,E.A.,Tap,W.D., D'Angelo,S.P.,Grupp,S.A., Holdich,T.,Binder-Scholl,G.,Jakobsen,B.K.,Odunsi,K.,Rapoport, A.andMackall,C. (2015) GeneticallyengineeredNY-ESO-1specificTcellsinHLA-A201+ patientswithadvanced cancers.J ClinOncol.33, suppl abstrTPS3102 2. Beatty,G.L., O'Hara, M.H., Nelson,A.M.,McGarvey,M., Torigian,D.A.,Lacey,S.F., Melenhorst,J.J.,Levine,B.,Plesa,G.andJune,C.H., (2015) Safetyandantitumoractivityof chimericantigenreceptormodifiedTcellsinpatientswithchemotherapyrefractory metastaticpancreaticcancer. J ClinOncol.33, suppl abstr TPS3007 3. Johnson,R.S., Walker,A.I.andWard,S.J.(2009) Cancer vaccines:will we everlearn? Expert RevAnticancerTher. 9(1), 67-74 4. MACS Miltenyi Biotec.(Accessed2015) CliniMACSProdigy速SystemMasteringthe complexityof cell processing. 5. Hagop M., Kantarjian,H.M., Fojo,T., Mathisen,M and Zwelling,L.A.(2013) Cancer Drugs in the UnitedStates:JustumPretium The JustPrice.AmSoc ClinOncol.31, 28, 3600-3604 6. Howard,D.H., Bach P.B.,BerndtE.R. and Conti,R.M.(2015) Pricinginthe Marketfor AnticancerDrugs.National Bureauof EconomicResearch,WorkingPaper20867, http://www.nber.org/papers/w20867 (accessed2015)

More Related Content

Commercialisation of Cellar Therapies for Cancer (final)

  • 1. Commercialisation of Cellular Therapies Anthony Walker, PhD and Robert Johnson Abstract Successful commercialisationof acell therapyrequiresmore thanprovingsafetyandefficacytothe regulators. The inherentcomplexityof cellularproductsdelivers particularmanufacturing,logistical and reimbursementhurdlesthatthreatencommercial viabilityforanytherapywithalessthan spectacularclinical profile thattrulychangesthe standardof care. Thisisparticularly acute for autologouscell therapieswhere patientsreceive bespoketreatmentsmanufacturedfromasample of theirowncellsandwhere economiesof scale,whichplayanimportantrole incontainingthe productioncostsfor small molecule andantibodytherapeutics,are highlylimited.Nevertheless,the promise of game-changingefficacy,asexemplifiedbyveryhighlevelsof complete responsesin refractoryhaematological malignancies,hasattractedcapital investmentsonavastscale,and the attendantpace of technologydevelopmentprovidespromisingindicatorsforfuture clinical and commercial success. Keywords:cellulartherapy,CAR-T,solidtumours,immunotherapy,oncology The CAR-T Landscape Clinical trialshave revealedspectacularefficacyinrefractoryleukaemiapatientsusinganti-CD19 chimericantigenreceptor(CAR) adoptiveT-cell therapy,witha92% complete remissionrate in39 patientswith Acute LymphocyticLeukaemia(ALL).These are transformative resultswhichhelptip the balance inthe waragainst cancer:if the complete responsesare durable,the technologyraises the previouslyunthinkablepossibilitythatlate stage cancercan be cured. These andotherdata have ignitedinvestorinterestinthe biotechsectorandthisfieldinparticular. Start-upcompanieshave raisedunprecedentedamountsof private capital,subsequentlytapped equallyimpressive sumsinIPOsand have gone onto achieve eye-wateringlyhighvaluationsonthe publicmarkets.Suchisthe appetite thatimportantinvestmentcriteria, forexample robustandclear IP have beensomewhatforgiven,althoughthishasbeenshiftingasthe sectorrapidlymatures. Big pharmahas alsodivedintothe field;inAugust2012 Novartisenteredintoanalliance withthe Universityof Pennsylvania(UPenn) onthe backof nascentclinical datawithU PennsCD19 construct.Novartisagreedtofunda CenterforAdvancedCellularTherapies(CACT) onthe U Penn campusin Philadelphia,contributing$20m.At the time of the deal,thiswasan extraordinary venture forNovartiswhichmaintainedastrongfocuson small moleculeandmonoclonal targeted therapies;throwingitsweightbehindanautologouscelltherapywasaradical departure forits typicallyconservative approachtodeal-making.The companyislookingtoaggressivelymaintainits leadershipinthe fieldthroughbothin-houseresearchandcontinueddeal-making. In March 2013 Celgene andBluebirdBioannouncedaglobal strategiccollaborationtoadvance gene therapyinoncology,withafocus on CAR-Technology.Financialtermsof the agreementinclude an upfrontpaymentandup to $225 millionperproductinpotential optionfeesandclinicaland regulatorymilestones. In June 2014 PfizerandCellectisenteredintoaglobal strategiccancerimmunotherapycollaboration to developCAR-Ttechnologies.PfizergainedaccesstoCellectisallogeneicapproachandPfizeris
  • 2. able to select15 targets.Cellectisreceivedanupfrontpaymentof $80m and milestonepaymentsof up to $185m. In January2015, AmgenandKite announcedastrategiccancer immunotherapycollaboration, combiningAmgensoncologytargetsandKitesCAR-Tplatformtodevelopnewtherapeutics.Kite received$80m upfrontfromAmgenandiseligibleforupto $525m in milestonesperprogram. Celgene redoubleditscommitmenttothe fieldinJune 2015, announcinga ten-yearcollaboration withJuno(albeitaftersignificantlynarrowingitscollaborationwithBluebird).Celgenegained optionstocommercialize Junosprogramsoutside NorthAmericaandmade aninvestmentof $1bn. However,the technologyfacesbigquestionstolive uptostratosphericexpectations,withatleast twomajor issuesfacingthe field.First,canthese technologiesbe commercialisedatreasonablecost (CAR-Ttechnologiesuse autologouscells,amajorlogistical challenge),andsecond,canthe technologybe appliedtocancersbeyondleukaemia.A technologythatdeliversequallyimpressive resultsina range of solid tumours wouldsurelymarkamilestoneinthe historyof medicine. Addressing Solid Tumours Althoughsuccessfultreatmentof advancedhaematological malignancieswouldrepresentamajor clinical win,commercial success relies onconqueringsolidtumourswhichrepresentover90% of the oncologysector.There isa dearthof clinical data;the mostimpressive todate include: A 60% response rate intensynovial sarcomapatientstreatedwithNY-ESO-1SpecificT-Cells producedbyAdaptimmune,withadurabilityof responserangingfromtwotonine months [1]; The U Penngrouptreatedpatientswithpancreaticductal adenocarcinoma(PDAC) with CAR-T cellsrecognisingmesothelin,anantigenoverexpressedonPDACcells.Twoof six patients experiencedstable disease byRECIST1.1 withdisease control off-therapyseeninone patientfor > 4 months.In one patient,abnormal 18FDG avidityseeninlivermetastasesatbaselinewasno longerdetectedat1 monthaftertherapy [2]. Whilstprovidingencouragingpreliminarysignsof efficacy,these resultsfall well shortof the possibly un reasonable game-changing benchmarksthatmanyinvestorsandsome BigPharmahave come to expectfromadoptive cellularimmunotherapy. Major technical challengesremain, notleasthomingof sufficientengineeredTcellstothe tumour site(s) toprovide anadequate efficacy:toxicity ratioforefficacy.Beyondthis,the microenvironment of a solidtumourmayprove to be far more hostile to CAR-Tcellsthancirculatingtumours,and inductionof anergy maybecome anefficacy-limitingfactorregardlessof CAR-Tcell persistence whichinitself maybe an issue.Finally,the emergence of CAR-Tresistantdisease throughantigen escape lossmechanismsalreadyseeninstudiesof cancervaccines [3] may limitthe durabilityof any responsesto CAR-Ttherapyand,as discussedbelow,thiscouldseverelyrestrictreimbursementfor thismodality. Manufacturing & Logistical Hurdles Most CAR-Tand TCR-engineeredTcells are currentlymade bya cumbersome andbespoke process involving: LeukapheresistoextractT cellsfroma cancer patientwhoisconnectedbytwointravenous tubesto an apheresismachineforseveral hours.Thisisnotcomfortable forthe patient,incursa
  • 3. substantial cost,andultimately,large-scaleadoptionof autologous CAR-Ttherapymaybecome rate limitedbyavailabilityof apheresiscapacity; Activationandtransductionof T cells.The original CAR-Tprocessesusedanti-CD3anti-CD28 DynabeadsforT cell activation;Novartismade anearlymove tosecure exclusive commercial access to thisreagentforCAR-Tproductiontoreinforce itsproprietarypositioninthe field.Since then,alternativestoDynabeadshave beensuccessfullyimplementedandthisnolonger representsacommercial hurdle.Transductionisusuallybyretroviral orlentiviral vectors, althoughnon-viral systemsare alsoused; Expansionof transducedTcellsoveran approximatelytwo-weekperiodinacytokine (typically IL-2) supplementedtissueculture medium; Washingand concentratingthe Tcellspriorto administration.For CAR-Tproducts made at central facilitiesandtransportedtoremote treatmentcentres,cryopreservationprotocolshave beendeveloped; QC release assaysare conductedforeachbatch of CAR-Tproduct. The entire processhasto be conductedunderenvironmentallycontrolledGMP compliantconditions whichare expensivetomaintainandrun.Aseach CAR-Tproduct ismade fromstartingmaterials(T cells) fromthe patienttobe treated,there are nosubstantial economiesof scale:increasing manufacturingcapacityrequiresthe additionof extraproductionpodsorboothsinparallel.The earliestproductionprocesseswere lab-based,fullymanual andinvolvedopentissueculture manipulationsinbiological safetycabinetsand/orisolators:althoughworkable forveryearlystage clinical research,thisapproachisclearlynotappropriate forcommercial purposes.Anindustrial grade manufacturingprocessneedstobe conductedinsealedsystems(zeroopenmanipulations) and shouldbe automatedasfar as possible,primarilytoreduce the humanvariabilityfroma complex processbutalsotoreduce costs. A fullysealedsystemcouldarguablybe situatedinaGrade D cleanroom(EU cGMP classifications),althoughindividual regulatoryagencieswill scrutinisethisin great detail andmayrequire Grade C conditionswhichincreasescomplexityandcost.Atpresent, the onlycommerciallyavailablesystemthatmeetsthe criteriaof a closedsystemautomated manufacturingdevice isthe Miltenyi CliniMACSProdigy [4]. Eventhen, noteveryprocesscanbe implementedonthatsystem(forexample, itappearsbettersuited tolentiviral transductionthan retroviral,andthisdependsonthe detailsof each specificprocesswhichwill notbe discussedhere). There are twoschoolsof thoughtregarding the locationof CAR-Tmanufacturing.The BigPharma industrial mind-setfavourslarge,centralisedfacilitiescapableof processingTcellsfrommany thousandsof patientsperyear.The alternative isadecentralisedmodel withmanufacturingco- locatedwiththe specialisttreatmentcentres(e.g.universityhospitals,majorcancertreatment centres).Theoretical modelsof the scalingof anautologousproductionmodulecanshow thatthere are optimumsize thresholdsforaproductionmodule atdifferentdose numbersperyear. Economiesof scale are non-linear:i.e. there are scenarioswheretwosmallproductionmodulesin parallel,eachcapable of producingsay500 dosesperyear,are more efficientthanamedium module scalesfor1,000 dosesperyear.These considerationsclearlyplayalarge role inthe decision betweenthe twomodels(andindeed the degreeof decentralisation;i.e.whatwouldthe optimum numberof sitesbe inthe USA; 25, 50, 100?). Top level pros andconsof eachmodel include are listedinTable 1.
  • 4. Pricing, Reimbursement & Market Access Whichevermanufacturingmodel ischosen,manufacturingcostsforautologous CAR-Tcell therapies are likelytobe inthe range of $25-35,000 perpatient,even aftermaximumprocessefficiencieshave beenexploited.Note thatthisisjustthe cost of manufacturingthe therapy,andinfact a course of treatmentincurssubstantial additional costsforpatientpreparation(mostcurrentprotocolscall for lymphodepletionpriorto CAR-Tadministration,anexpensive andtime consumingin-patient procedure), CAR-Tadministrationandfollow upcare costs(notablythe resourcesrequiredfor monitoringandtreatingthe severe sideeffectse.g.cytokine stormoftenassociatedwithCAR-T therapy). Total treatmentcostsmaytherefore be ashighas those for autologous bone marrow transplantation (currentlyca.$360,000 in the USA) plusthe price of the CAR-T therapy. In the immediate aftermathof the Novartisdeal with UPenn,financialanalystswereestimatingthat CAR-Ttherapieswouldbe pricedat$250,000 per patientandpossiblyhigher. AsnoCAR-Tproducts have beenlaunchedontothe marketyet, future pricingremainsspeculative,butmany analystsare now basingforecasts onprice ranges between$150,000 and$300,000 per patient.Eventhe lower endof thisrange isabove current annualisedmedicationcostsfornewly-launchedcancerdrugs. At firstglance,this appearstoflyin the face of the growingdebate aboutthe JustumPretium,orJust Price,of cancer drugs [5], voluble clinicianresistancetoperceivedexcessive drugpricing(arguably startedinOctober2012 whenphysiciansatthe Memorial Sloan-KetteringCancerCenterannounced inthe NewYorkTimesthat theirhospital wouldntbe usingZaltrap,anew VEGF-targetedcancer treatmentformetastaticcolorectal cancer) andthe increasinguse of HealthTechnologyAssessment (HTA) to restrictthe use of highpricedtreatments. A recentstudyassessinganoriginal dataset of 58 anticancerdrugsapprovedbetween1995 and 2013 foundthatlaunchprices,adjustedforinflation and drugssurvival benefits,increasedby10%,or about $8,500, peryear[6]. Thus,in 2004 bevacizumabwas launched forpatientswithlate stage colorectal cancerata price of $50,000 per treatmentepisodeandwasassociatedwithanincremental increase inoverall survival of five months.Sevenyearslaterin2011, ipilimumabwasapprovedfortreatmentof melanoma,associated withan incremental increase inlifeexpectancyof fourmonths,andlaunchedata price of $120,000 pertreatmentepisode.Itisclearlyuntenableforthistrendtocontinue indefinitelyinanalready budget-constrainedhealthcare environment. The keyquestiontherefore is:willpayersbe preparedtofund CAR-Ttreatmentatthese levels?This iskeyto understandingthe commercial prospectsforthe technologyasawhole.The answersimply isthat it will dependonthe level of clinical efficacythatcan be achievedwiththis new modality.In the most optimistic(but probablyunrealistic) scenario,asingle CAR-Tadministrationwouldinduce longtermremission,sayfive totenyears.Usingthe UKs NICEbenchmarkof 贈30,000 ($47,000) per quality-adjustedlife year,thiswould amounttoa maximumreimbursableprice of $235-470,000. Thiswouldhave to be adjusteddownwardstoaccountforthe fact thatcancer patientsqualityof life islessthanperfect,butthatstill mayallow aprice for CAR-Ttreatmentof $118-235,000. Allowingforthe observationthatNICEsapproachisamongstthe moststringentinmainstream pharmaceutical markets,andUKpricesfor pharmaceuticalstendtobe significantlylowerthaninthe US (the mainmarket) and othercountries,itispossiblethatthe HTA-acceptedpricesforCAR-T therapywill be above thisrange andtherefore inline withcurrentanalystprojections.However,any shortfall inefficacybelowadurable five-yearremissionwouldnecessarilyreducethese calculated price ranges.Thismay perhapsbe an overlysophisticatedwayof sayingthat CAR-Ttherapieswill be paidfor onlyif theyworkinthe clinic,butthe seriousconclusionisthatthe benchmarkfor commercially-viable efficacyisgoingtobe muchhigherthanfor lesscomplex modernoncology
  • 5. drugssuch as monoclonal antibodies. Unless CAR-Ttherapy, initscurrentpatient-specific autologousapproach,caninduce longtermdurable remissionsinahighproportionof treated patients,the chancesof commercial successare low. New Technology to Drive Commercial Success A numberof technologies onthe horizonappearsettogreatlyimprove the clinical andcommercial prospects foradoptive cellularimmunotherapy. Some of these approachesaimatenhancingthe efficacyof CAR-Tcells,forexample byboostingthe homingof infusedcellstotumoursites,the lack of whichisbelievedtobe amajor reasonwhyefficacyof CAR-Tcellsinsolidtumoursissignificantly belowthatincirculatingtumours.There are alsoapproachesto enhancingthe endurance of CAR-T cellsnotjustin termsof the lengthof durationinthe circulation, butalsothe retentionof anactive, non-anergised phenotype.Additionally,some researchersare assessingmethodstobuildin defencesagainsttumourmechanismsthat evade ordisable the immune system;suchanti-missile missiles mayhave alot of mileage inboosting CAR-Tefficacy. Anothersetof approachesfocuson the safetyof CAR-Ttherapies.Whilstthereissome evidence that cytokine releasesyndromeisatleastassociated,if notcorrelated withclinical efficacy,there may be methodstobluntthe severityof the reactionwithouttoomuchof an efficacyhit.Reducing the needto treatthe side effectsof CAR-Ttherapyinan intensive care unitwouldclearlybe better for patients,more practical fortreatmentcentresandmuchmore attractive to payers. Perhapsthe mostfundamental approachtoimprove cellularimmunotherapywill be totransformit froma bespoke,patientspecificprocesstoa mass-producedproduct. Thiswouldreduce leadtimes for treatment(the productwouldbe available off-the-shelf), have aninherentlylowerproduction cost (allowingforreductionsinpricingwithoutoverlysacrificingprofitmargins)andwouldbe more inline withthe model of productdistributionbythe pharmaceutical industry. The firststeptoward thisisthe use of allogeneicTcellsfromhealthydonors,forexample gene editedtoeliminate the endogenousTcell receptorandthuspreventacute rejectionbythe recipient,andwhere one donor can provide enoughcellstotreatmultiplepatients.Ultimately,we foresee thatthisapproachwill extendtoimmortalisedTcell linesthatcanbe culturedinbulkon an industrial scale,eliminatingthe needforT cell donorsaltogether.Animmortalised,allogeneicmassproduced CAR-Tcell capable of inducingdurable (five yearsormore) remissionsremainsapipe dream, butatleastis onthe radar for a numberof well-resourcedresearchgroupsandthusstandsa reasonable chance of becominga reality.
  • 6. Table 1: Toplevel pros andcons of manufacturingmodels Pro Con Centralised - Economiesof scale (albeitlimited) - Extremelycomplex logistics increasesriskof production failures/errors - Higherriskof productionoutages inthe eventof plant contamination,etc. Decentralised - Enablesuse of fresh,non- cryopreservedproductwhichis likelytobe inherentlysuperior - Maximiseslocal clinicianbuy-into the therapy - Potentiallyharnessesexisting investmentsincleanroomsat majorcancer centres - Needtodemonstrate comparabilityof manufacturingat each site:complex,time consuming,costly - Validatinganyprocesschange at each site multipliesthe efforts requiredvsa centralisedsetting
  • 7. References 1. Merchant, S.,Cristea,M.C., Stadtmauer,E.A.,Tap,W.D., D'Angelo,S.P.,Grupp,S.A., Holdich,T.,Binder-Scholl,G.,Jakobsen,B.K.,Odunsi,K.,Rapoport, A.andMackall,C. (2015) GeneticallyengineeredNY-ESO-1specificTcellsinHLA-A201+ patientswithadvanced cancers.J ClinOncol.33, suppl abstrTPS3102 2. Beatty,G.L., O'Hara, M.H., Nelson,A.M.,McGarvey,M., Torigian,D.A.,Lacey,S.F., Melenhorst,J.J.,Levine,B.,Plesa,G.andJune,C.H., (2015) Safetyandantitumoractivityof chimericantigenreceptormodifiedTcellsinpatientswithchemotherapyrefractory metastaticpancreaticcancer. J ClinOncol.33, suppl abstr TPS3007 3. Johnson,R.S., Walker,A.I.andWard,S.J.(2009) Cancer vaccines:will we everlearn? Expert RevAnticancerTher. 9(1), 67-74 4. MACS Miltenyi Biotec.(Accessed2015) CliniMACSProdigy速SystemMasteringthe complexityof cell processing. 5. Hagop M., Kantarjian,H.M., Fojo,T., Mathisen,M and Zwelling,L.A.(2013) Cancer Drugs in the UnitedStates:JustumPretium The JustPrice.AmSoc ClinOncol.31, 28, 3600-3604 6. Howard,D.H., Bach P.B.,BerndtE.R. and Conti,R.M.(2015) Pricinginthe Marketfor AnticancerDrugs.National Bureauof EconomicResearch,WorkingPaper20867, http://www.nber.org/papers/w20867 (accessed2015)