Efficacy trials of ZIKV Vaccines: endpoints, trial design, site selection

Transcription

1 Efficacy trials f ZIKV Vaccines: endpints, trial design, site selectin WHO Wrkshp Meeting Reprt June 1-2, 2017 Htel Ryal Mantel, Geneva

2 Table f cntents TABLE OF CONTENTS 1 1. INTRODUCTION 2 2. CONTEXT 3 3. TRIAL DESIGN CONSIDERATIONS CLINICAL ENDPOINTS SELECTION CASE DEFINITION CASE ASCERTAINMENT IMMUNOLOGICAL ENDPOINTS TRIAL DESIGN ELEMENTS 8 4. ESTABLISHING A TRANSPARENT FRAMEWORK FOR SELECTING VACCINES TO BE EVALUATED IN PHASE 2B/PHASE 3 TRIALS NEXT STEPS 12 1

3 1. Intrductin WHO is wrking n the R&D Blueprint. It is a glbal strategy and preparedness plan that allws the rapid activatin f R&D activities during epidemics. Its aim is t fast-track the availability f effective tests, vaccines and medicines that can be used t save lives and avert large scale crisis. With WHO as cnvener, the brad glbal calitin f experts wh have cntributed t the Blueprint cme frm several medical, scientific and regulatry backgrunds. WHO Member States welcmed the develpment f the Blueprint at the Wrld Health Assembly in May Amng the varius activities, the Blueprint is fstering the develpment f innvative study designs fr pririty pathgens, starting with vaccines. On 1 February 2016, the Wrld Health Organizatin (WHO) declared a Public Health Emergency f Internatinal Cncern (PHEIC) fllwing the unusual incidence f Zika virus (ZIKV) disease and the strng assciatin, in time and place in Latin America and the Caribbean, between ZIKV infectin and a rise in cases f cngenital malfrmatins, particularly micrcephaly, and neurlgical cmplicatins. Fllwing the declaratin f PHEIC, the develpment a ZIKV vaccine assumed high pririty. Vaccine clinical develpment was refined as understanding f pathgenesis and disease dynamics evlved, and new diagnstic and serlgical assays are being develped. The WHO vaccine pipeline tracker currently registers 45 ZIKV vaccine candidates under develpment, by private and public-sectr develpers. Six f these candidates have entered clinical trials. Under the Blueprint Plan f Actin, WHO has led a series f initiatives t maintain cntinuus dialgue between develpers, regulatrs and public health experts t identify hw best t achieve rapid, rbust, safe, and evidence-based licensing f ZIKV vaccines. In March 2017, the WHO Glbal Crdinatin Mechanism fr Research and Develpment reviewed the ZIKV vaccine develpment pipeline, and called fr imprved cllabratin and crdinatin t accelerate ZIKV vaccine evaluatin. On June 1-2, 2017, the WHO cnvened a grup f abut 30 experts in epidemilgy, regulatry, preclinical and clinical vaccine trials, and mathematical mdelling, in a wrkshp n planning fr Zika vaccine efficacy trials. The wrkshp aimed t define generic principles n hw t best design, cnduct and analyze vaccine trials against ZIKV, based n the available scientific evidence as well as n lessns learned frm evaluating ther flavivirus vaccines. The wrkshp builds n WHO effrts in Zika vaccine R&D ( In particular, we kindly invite readers t familiarize themselves with the Zika vaccine Target Prduct Prfile fr emergency and rutine uses as well as the WHO Meeting Reprt n cnsideratins fr regulatry expectatins f Zika virus vaccines fr use during an emergency ( ). Participants reviewed available evidence, identified and discussed methdlgical ptins t evaluate vaccines, regardless f vaccine prducts, and agreed n sme preliminary recmmendatins. It was recgnised bth that the preliminary recmmendatins are likely t evlve as new evidence is generated and als, they must be tailred t the scial and cultural cntext f affected cmmunities. 2

4 2. Cntext There is still n-ging ZIKV transmissin in Latin America with hetergeneus transmissin patterns in time and space. After its intrductin in Brazil, ZIKV spread in many parts f Latin America and the Caribbean like wild fire, where the vectr Aedes is present. Transmissin exhibited hetergeneity at different scales. At the glbal level transmissin patterns are primarily explained by the seasnal and gegraphical variability in vectr density. At the reginal level, surveillance data in Brazil, revealed variable incidences f micrcephaly, suggesting ccurrence f distinct lcalized utbreaks f ZIKV infectin. Hwever, the interpretatin f these data is difficult because the availability f surveillance and mnitring data varies between areas and such data as are available may nt have been based n standardized definitins, instruments and measures acrss regins. At the lcal level, hetergeneity in transmissin may arise frm differences in the extent t which peple prtect themselves, r are prtected, frm infectin. Several knwledge gaps remain: the rle f factrs such as age, sci-ecnmic status, and ther demgraphic factrs as risk and prtective factrs requires further investigatin t identify at-risk ppulatins and t understand what might ptentially influence transmissin rates and affect the feasibility f evaluating a vaccine. In additin, there is emerging evidence that prir immunity t dengue may affect ZIKV susceptibility, and ptentially impact n transmissin, and the vaccine perfrmance may be affected by whether r nt recipients have been infected with, r received vaccines against, ther flaviviruses. The hetergeneity in transmissin and lack f standardizatin f methds used t determine ser-cnversin results cmplicates the transferability f findings t different regins and highlight the need fr prspective use f generic prtcls. Ser-surveys culd be used t prvide infrmatin abut prir transmissin f ZIKV and the prprtin f ppulatins that remain susceptible, and sequential ser-surveys might be used t estimate transmissin rates. Hwever, interpretatin f these wuld be facilitated by standardisatin f methds and the use f generic prtcls, as well as the use f serlgic tests that can distinguish ZIKV infectin frm thse due t ther flaviviruses (and thse vaccinated against ther Flaviviruses, as well, in the future, any previusly vaccinated against ZIKV). There are currently n licensed ZIKV vaccines but there are 45 vaccine candidates registered in the WHO vaccine pipeline tracker. WHO has develped a Target Prduct Prfile (TPP) that sets public health prduct preferences fr emergency and rutine ZIKV vaccine use. The TPP is designed t prvide aspiratinal guidance t vaccine develpers, and is infrmed by regulatry expectatins and by technlgical feasibility. As there are many vaccines candidates (thugh nt all in the pipeline are expected t prgress t clinical testing), the activities cnducted under the Blueprint are aiming t explre 3

5 the best trial strategies and designs t test ne r mre prmising candidate vaccines. In a cntext where there is n established immune marker f prtectin against ZIKV, clinical trials t evaluate vaccine efficacy will be necessary, and these may als help define a crrelate f prtectin. It is hped that that the availability f a transparent framewrk t review varius candidates attributes t help infrm the selectin f thse t be taken int clinical trials wuld cntribute t ensure resurces are utilized mst efficiently, aimed at evaluating and licensing efficacius vaccines. 3. Trial design cnsideratins 3.1 Clinical endpints selectin Critical issues that were cnsidered are: (i) Can we use ZCS as an endpint? (ii) is ZCS a result f clinical Zika r can it be caused by subclinical ( asymptmatic) zika? (iii) is it pssible that a zika vaccine has different prtectin fr clinical and sub clinical Zika? The demnstratin f benefit based n a clinical endpint is the ptimal way t evaluate a ZIKV vaccine but, if this is nt feasible, ther appraches may be necessary. Licensure f vaccines generally requires the demnstratin f benefit based n a clinical endpint r based n a scientifically well-established marker f prtectin, using evidence generated by well-cntrlled clinical studies. T reduce the time lag f access t vaccines in emergency cnditins, especially when efficacy trials may be impssible t cnduct, regulatry agencies have develped alternative regulatry pathways. Fr example, in additin t the traditinal apprval pathway, the US Fd and Drug Administratin has delineated tw ther ptential pathways t license a vaccine, requiring varying level f scientific, clinical and pre-clinical evidence that supprt the benefits f the vaccine in humans. Accelerated apprval is based n the demnstratin f a surrgate f prtectin thugh well-cntrlled clinical studies that are reasnably likely predict clinical benefit. The US FDA animal rule is based n the demnstratin f an immune marker f prtectin in animal mdels that is reasnably likely predict clinical benefits in humans. Bth accelerated apprval and animal rule appraches require pstlicensure studies t verify clinical benefit and safety. There is grwing understanding f ZIKV pathgenesis in the general ppulatin, pregnant wmen, and nenates, but knwledge gaps persist. Clinical endpints that may be used in vaccine trials range frm rare and mre severe cmplicatins f ZIKV infectin t mre cmmn and mild clinical manifestatins. A causal assciatin between ZIKV infectin in pregnant wmen and micrcephaly in nenates has been demnstrated in different settings. The spectrum f cnditins cmprising Zika cngenital syndrme (ZCS) is expanding based n clinical studies, and micrcephaly is likely t cmprise nly a relatively small prprtin f all cngenital abnrmalities assciated with ZIKV infectin in pregnancy (1-3% fr micrcephaly fllwing ZIKV infectin in pregnancy and mre frequent rates fr ZCS). 4

6 Neurlgical cmplicatins may ccur fllwing ZIKV infectin, thugh are less frequent than ZCS. The natural histry f ZIKV-induced Guillain-Barré Syndrme (GBS) remains unclear. GBS is t rare event fr the incidence t be measured reliably in prspective chrt studies, unless these are very large. Fllwing ZIKV infectin it has been estimated that the incidence f GBS is f the rder f % and that fr any neurlgical cmplicatins abut %). Emerging evidence suggests that ZIKV infectin is a trigger t develpment f GBS. It is unclear whether ZIKV-triggered GBS is mediated either by direct viral effect r by ZIKV antibdies the latter wuld raise safety cncerns in ZIKV vaccine develpment. Mre research is needed t better understand the full clinical spectrum f ZIKV illness and its clinical manifestatins and pathgenesis. WHO is develping generic prtcls t facilitate the synthesis f data generated by varius grups and the interpretatin f results. 3.2 Case definitin ZCS and neurlgical cmplicatins d nt seem feasible primary endpints fr vaccine efficacy trials but may be investigated in studies cnducted pst-licensure. The primary public health burden f ZIKV infectin is assciated with ZCS and neurlgical cmplicatins. Therefre, ZCS and neurlgical cmplicatins might seem the apprpriate chice fr clinical endpints, and wmen f childbearing age the primary target ppulatin f interest fr ZIKV vaccine evaluatin, in line with the target ppulatin indicated in the Zika vaccine TPP. Hwever, ZCS and neurlgical cmplicatins ccur at t lw a frequency t be chsen as primary endpints in Phase 2b r 3 vaccine efficacy trials. The chice f mre frequent clinical events as a primary utcme measure fr vaccine efficacy trials is likely t be necessary fr trials f feasible size. The justificatin f a mild, mre cmmn, endpint as the primary endpint in vaccine trials wuld be predicated n the assumptin that the benefit f the vaccine n the selected mild endpint is reasnably likely t predict clinical benefit fr severe cmplicatins (e.g. ZCS in nenates and neurlgical cmplicatins), and prvisin will be made t verify the benefit against mre severe disease a psteriri. ZIKV infectin as a ptential endpint fr vaccine trials. Abut 80% f ZIKV infectins are asymptmatic. The rle f asymptmatic infectins in pregnant wmen in causing ZCS is unclear, hwever ZCS has ccurred fllwing such asymptmatic infectin. It is als unclear if viral lad titres fllwing infectin are assciated with clinical illness, ZCS r neurlgical cmplicatins. A pssible endpint fr clinical trials wuld be ZIKV infectin (whether symptmatic r nt). Hwever, detecting asymptmatic ZIKV infectins is nt straightfrward. Detecting an active infectin, with the presence f ZIKV in bld r urine, requires very frequent cllectins f bld and urine pst-vaccinatin, as current assays can nly detect virus fr a shrt perid (1 r 2 weeks fllwing infectin, which raises cncerns f acceptability t trial participants). RT-PCR prvides the gld standard methd t detect the viral presence in bld, r urine, and t estimate viral lad. Sampling bth bld and urine can help maximize RNA detectin sensitivity. 5

7 Evidence f infectin may als be gained thrugh paired serlgical measurements, prvided that a suitable serlgical assay fr ZIKV infectin is develped, that can distinguish ZIKV infectin frm infectin by ther flaviviruses and frm serlgical respnses prduced by the ZIKV vaccine under evaluatin. Hwever, t ascertain the timing f any infectin, relatively frequent bld samples wuld have t be taken during a vaccine trial. Virlgically-cnfirmed ZIKV illness is a cnvenient and feasible primary endpint in a vaccine efficacy trial. A challenge with using clinical disease as the primary endpint in vaccine trials is that ZIKV illness is ften assciated with mild symptms, which raises challenges fr case detectin. Cases f illness due t ZIKV infectin may be missed unless there is an intensive trial surveillance system (invlving hspitals, clinics, and frequent cntact with participants t elicit symptms which may be due t ZIKV infectin). A standardized clinical case definitin is essential t facilitate cmparing and cmbining infrmatin frm different studies. Lessns learned frm dengue vaccine trials als underscre the need fr active surveillance in relatin t fairly mild symptms. It is als pssible that the ZIKV clinical case definitin might evlve ver time, as mre clinical features are identified which are assciated with the infectin. Mre research is needed t better understand the natural histry f ZIKV infectins, including the clinical spectrum f ZIKV illness, the incubatin perid and duratin f symptms. ZIKV symptms (e.g. rash, fever) are similar t thse assciated with ther flavivirus infectins, such as dengue r chikungunya, which might be c-circulating, and therefre definitive diagnsis f disease due t ZIKV requires timely cnfirmatin by labratry testing f apprpriate tissue samples. 3.3 Case ascertainment RT-PCR is the currently prpsed gld standard fr cnfirming the presence f ZIKV RT-PCR testing f serum, urine, saliva, whle bld, r ther tissues is a specific test fr ZIKV infectin. Hwever, the time windw fr sample cllectin fllwing infectin t cnfirm diagnsis is shrt, especially fr bld (within 4-7 days after symptms nset). Cmbining results frm different tissue samples t maximize RNA detectin can be used t increase sensitivity. In additin, RT-PCR can be used t measure viral lad in tissue samples. Case ascertainment f ZIKV infectin thrugh scheduled sample cllectin Prvided that the vaccine des nt express NS1 prtein, NS1-partial_E IgG and/r NS1- partial_e-igm, ELISA assays culd be used t ascertain sercnversin events during the trial fllw-up perid, the timing f such events being determined by the frequency f bld sampling. Testing fr IgM antibdies with NS1-partial_E-IgM-based ELISA assays may be used t cnfirm ZIKV disease acquired recently, with sme uncertainties arund the timing f infectin. Because f ZIKV trpism in a variety f tissues, IgM may be detected in bld fr lnger than ther tissues. 6

8 3.4 Immunlgical endpints Neutralizing antibdies titres induced by vaccinatin and measured by PRNT may prvide a surrgate f prtectin fr ZIKV vaccines as reasnably likely t predict clinical benefit. Neutralizing antibdies are well-established markers f prtectin fr vaccines targeting sme ther flaviviruses (e.g. yellw fever, Japanese Encephalitis, but nt dengue). The plaque reductin neutralizatin test (PRNT) prvides a gld standard in measuring neutralizing antibdies and serves t define prtective titres fr sme ther flavivirus vaccines. Hwever, the PRNT assay requires bisafety equipment, is labur-intensive and may take a week t cnduct. Als, PRNT cannt distinguish between wild-type and vaccine induced immunity. Passive antibdies transfer studies shw that purified IgG frm vaccinated nn-human primates (NHP) prtects mice and NHP (absence f viraemia) fllwing ZIKV challenge. Animal mdels shws significant variability in neutralizing antibdies titers depending n what prtectin is measured against. Fr example, it has been reprted in NHP challenge mdels that neutralizing antibdies titers f 1:100 prevent viraemia, but that 1:5000 is required fr sterilizing immunity (defined as n increase in neutralizing antibdies fllwing ZIKV challenge). In additin, the titer f neutralizing antibdy required fr prtectin may vary depending n the vaccine platfrm utilized t induce an immune respnse. Althugh neutralizing antibdies appear qualitatively similar t thse fr ther flaviviruses, the quantitative titers required fr ZIKV prtectin appear t be much higher in animal mdels cmpared t ther flaviviruses, which may be due t trpism with many tissues. While there are grwing indicatins n the feasibility f defining neutralizing antibdies as a crrelate f prtectin fr ZIKV in animal mdels, translatin f the results int humans remains uncertain. Lessns learned frm dengue and ther vaccine -preventable diseases underscre the ptential rle f cntrlled human infectin mdels (CHIM) t explre hw neutralizing antibdies crrelate with different levels f prtectin. Such studies n ZIKV may help screen vaccine candidates and infrm hw efficacy trials shuld be designed t investigate and validate a crrelate f prtectin. Hwever, it was nted that an ethics review cmmittee had recently been cnvened t cnsider CHIM fr Zika and had cncluded that, because f uncertain safety issues, it wuld be premature t prceed with a CHIM at this time. Standardized, validated assays, with agreed units f measurement will be critical t quantitate neutralizing antibdies, especially if lking at backgrund f multiple flaviviruses cncurrently. 7

9 3.5 Trial Design elements A trial ppulatin representative f the general ppulatin shuld be cnsidered Althugh the primary target ppulatin fr vaccinatin wuld be wmen f child bearing age (as specified in the WHO TPP), if clinical disease due t ZIKV infectin is selected as the primary endpint fr trials, then trials culd be cnducted in bth men and wmen. Hwever, clinical benefit t the primary target ppulatin, as described in the TPP, must be verified, prbably pst-licensure. Fr vaccine candidates that wuld be expected t have a favurable prfile fr pregnancy, pregnant wmen culd be included in the trial at sme stage. The need fr a multi-site apprach Given that predictin f where utbreaks f ZIKV infectin will ccur at specific times is prblematic, a multi-site apprach fr vaccine trials may be apprpriate t increase the chance f including grups with a high incidence f disease, as well as prviding an pprtunity t evaluate vaccine efficacy acrss different ppulatins. A duble-blind placeb-cntrlled individually-randmized trial is the ptimal design t evaluate the efficacy f a ZIKV vaccine candidate. Individual level f randmizatin is preferred - Individual level randmizatin is preferred t a cluster-randmised trial design because f the likely substantial variatin in the incidence ZIKV infectin frm area t area, which mitigates against a cluster-randmised design. With individual randmisatin multiple vaccines culd ptentially be tested simultaneusly in the same trial. Masking prcedures: placeb is preferred - As there is n existing ZIKV vaccine, a placeb-cntrlled trial wuld be ethically acceptable. A vaccine against anther disease, that the trial ppulatin wuld nt nrmally receive and that des nt affect the incidence f the primary and secndary endpints, might als be cnsidered. Hwever, assessment f the reactgenicity f the ZIKV vaccine may be hampered if the cmparatr vaccine is highly reactgenic (which might als cmprmise blinding). Define a statistical analysis plan a priri - A statistical analysis plan shuld be prepared prir t the start f a trial. This shuld include cnsideratin f any interim analyses, with specificatin f the circumstances in which the trial wuld be halted fr verwhelming efficacy r fr futility. The analysis wuld likely invlve cmbining data acrss sites. Adaptive trial designs are acceptable frm a regulatry perspective (fr example, drpping a prly perfrming vaccine early, if several are being tested in the same trial), but all g/n g decisins need t be established in advance. A multi-site trial requires standardizatin f cncepts (e.g. same prtcl, ne DSMB), instruments (e.g. labratry assays), and measures (e.g. generic prtcl t diagnse ZIKV illness). Sites in which n ZIKV cases ccurred wuld nly be used fr vaccine safety data. The duratin f fllw-up f participants may have t be extended if an expected ZIKV utbreak fails t materialise. One strategy wuld be t set up Phase 2 trial sites fr safety and immungenicity assessment in at-risk areas and switch a site t a Phase 2b/Phase 3 prtcl as sn as ZIKV transmissin is detected, accrding t a predefined criterin. Hwever, early detectin f lcalized ZIKV transmissin is difficult and it may take several 8

10 mnths t set up a trial at a given lcatin. It was nted that analysing msquit pls t detect ZIKV as a marker f an incipient utbreak may nt be a useful strategy, based n experience with West Nile virus. Furthermre, trial designs evaluating vaccines with multiple dse-regimen may be subject t an additinal delay befre being able t cunt cases accrding t a per-prtcl analysis, which may be a limitatin if the transmissin perid is shrt. Leveraging all available evidence t infrm site selectin and study design (e.g. trial simulatrs, ser-prevalence studies, chrt studies) Future patterns f transmissin f ZIKV in Latin American, the Caribbean and elsewhere are difficult t predict, but it is likely that lcalized transmissin will ccur in sme places, pssibly with lw level ZIKV endemicity. Circulatin f ZIKV in African and Asian cuntries has als been reprted. As ZIKV is transmitted by the same vectr as dengue, ZIKV infectin and dengue may fllw a similar gegraphic distributin. Mre research is needed t assess the distributin and virulence f ZIKV strains acrss cntinents. Ser-prevalence surveys and vectr mapping can help infrm site selectin fr vaccine trials, as it was the case fr dengue vaccine trials. Als, Zika and dengue share similar transmissin patterns, and previus htspts fr dengue may predict ZIKV transmissin. Three mdelling grups are making ZIKV prjectins fr 2017 t estimate where ZIKV infectin attack rates are likely t be ver 5% in the year, t help infrm selectin f sites fr Phase 2b r 3 vaccine trials. Prjectins will be validated using surveillance data. Fr 2018, ne f the mdels prjected that lcalized utbreaks are still likely t ccur in Latin America and are likely t be triggered by ppulatin mvements. Serprevalence results, frm crss-sectinal r chrt studies, wuld prvide a precius and cmplementary dataset at the ppulatin level t imprve the accuracy f prjectins and the rbustness f mdels, and, thus better infrm site selectin. Als, mdels wuld benefit frm integrating data n ther flaviviruses t better prject the distributin f at risk ppulatins. Finally, mst f the mdelling t date has fcussed n Latin America and there is a need t prject the incidence f ZIKV infectins in ther parts f the wrld. There is a need t cllect pre-vaccinatin bld samples in trials ZIKV efficacy trials shuld be designed t be able t take accunt f prir immunity. Lessns learned frm dengue vaccine efficacy trials underscre the need t cllect bld samples just befre vaccinatin frm all participants, and nt just frm a subset. A dried bld spt may be adequate and wuld likely increase acceptability. The questin f whether thse with a previus ZIKV infectin shuld be excluded frm efficacy trials was discussed. Ideally, the trial ppulatin wuld be naïve t ZIKV infectin if ZIKV infectin cnfers lasting immunity against a further infectin. Hwever, screening fr prir ZIKV infectin was deemed inefficient, unless a trial was cnducted in an area f high previus infectin. Furthermre, in ppulatin-based vaccinatin prgrammes, such screening wuld be unlikely t be feasible. Furthermre, it remains unclear whether r nt ZIKV infectin des prvide lifelng immunity. 9

11 It was agreed that prir immunity t ZIKV r any ther flavivirus shuld nt be an exclusin criteria fr participants in efficacy trials and that serstatus at baseline shuld be used fr stratified analysis. A specific assay, like PRNT r NS1-IgG ELISA, shuld be used t determine serstatus at baseline and be used t distinguish between prir infectins with different flaviviruses. Hwever, the ELISA assay needs further evaluatin and validatin, especially in ppulatins with high rates f secndary flavivirus infectins. PRNT wuld als be used t measure the increase in neutralizing antibdies fllwing vaccinatin, t better understand the ptential fr prtectin. Mrever, it shuld be nted that PRNT cannt distinguish between wild-type and vaccine induced immunity. 4. Establishing a transparent framewrk fr selecting vaccines t be evaluated in Phase 2b/Phase 3 trials Given the number f candidate vaccines under develpment and the challenges f identifying and establishing trial sites, it was discussed that there may be merit fr transparent and evidence based apprach fr selectin f candidate vaccines fr trials. Sme initial cnsideratins were discussed. There was a draft prpsal t cnsider tw categries f criteria: required and desirable. Regarding the required criteria the fllwing elements were prpsed: Pre-clinical efficacy: (i) demnstrate clse t 100% prtectin against viraemia in primate (human r NHP) mdel?; (with the caveat that CHIM has nt yet been fully develped). Such prtectin a muse mdel alne wuld prbably nt be cnsidered sufficient) Phase 1/Phase 2 clinical studies: data including: (i) Flavi-naïve and nnflavi-naïve subjects; (ii) immungenicity with greater f equal prtective levels than bserved in NHP challenge studies?; (iii) acceptable safety/reactgenicity regardless f prir flavivirus expsure Regarding the desired criteria, the fllwing elements were prpsed: General cncurrence with the elements nted in the WHO TPP: number f dses, length f schedule, suitability fr pregnant wmen, stability, duratin f immunity, etc. Evidence that prduct prductin can be scaled up t prduce sufficient GMP grade dses fr clinical evaluatin and beynd. Capabilities f manufacturer r future arrangements in place: Clinical trial infrastructure (efficacy trial experience, pharmacvigilance, manufacturing capacity) Cnsidering the abve a draft grid fr priritizatin was sketched. It is imprtant t nte that this is an early versin and additinal wrk and cnsultatins will be required. 10

12 Draft Grid fr priritizatin (this is an early versin with hypthetical infrmatin included fr illustratin purpses nly) Vaccine candidate attributes Prtectin against viremia (CHIM>NHP> mice>nne) Safety in Phase 1/2a Safety in target ppulatin Dse regimen (single vs multiple dses) Clinical trial expertise Scalability (manufacturing capability, Pharmacvigilanc e) Candidate Candidate Candidate Candidate Candidate

13 5. Next Steps A series f cllabrative steps t cntinue t advance the discussins were utlined and agreed upn. It is anticipated that the steps will be implemented in clse cllabratin with the wrkshp participants and ther experts in the cmmunity as apprpriate. a) Develping an anntated interactive generic prtcl fr ZIKV vaccine efficacy trials, based n preliminary design cnsensus. The generic prtcl will be develped with inputs frm all participants, and it will be published in the WHO website fr public cnsultatin by Fall It is anticipated that candidate vaccine develpers will be infrmed f the cnsultatin prcess and invited t prvide their perspective. b) Prmte data cllectin standardizatin and incentivize data sharing thrugh the WHO R&D bservatry. Ser-prevalence surveys and multi-site ser-cnversin chrts in different ppulatins, Aedes density infrmatin, and mdelling studies are crucial t better understand the distributin f susceptible ppulatin and t anticipate patterns f transmissin in the future, and t infrm the planning f vaccine evaluatin studies. Therefre, an effrt twards standardizatin f data definitin and cllectin, thrugh the develpment and use f generic WHO prtcls, is encuraged t btain a glbal picture f transmissin. c) Establish cllabratins with cuntries and research sites t leverage all available evidence. WHO will cntinue t prmte cllabratin with relevant grups, including industry, t re-analyze samples that may have been cllected fr ther purpses (such as ther clinical trials cnducted in ZIKV endemic regins where samples were cllected) t gain knwledge f ZIKV transmissin that may have therwise gne unnticed. WHO is cllabrating with cuntries t re-analyze natinal and reginal bld bank specimens. The infrmatin and results will be available thrugh the WHO R&D bservatry t ensure disseminatin f all pertinent infrmatin. d) Establishing a transparent framewrk fr selecting vaccines t be evaluated in Phase 2b/Phase 3 trials. There are 45 ZIKV vaccine candidates in varius stages f develpment and there wuld be benefit frm establishing a framewrk fr vaccine selectin t ensure resurces are utilized fr develpment f candidates that demnstrate the highest likelihd f success. Similarly t the TPP, this framewrk wuld set WHO preferences and wuld prvide aspiratinal guidance fr vaccines t be priritized fr Phase 2b/Phase 3 trials. Preferences wuld be established thrugh a list f required and desired attributes and criteria, including pre-clinical and early clinical evidence, as well as cmpliance with TPP features, and scalability f the prduct. Using the initial inputs at this cnsultatin, WHO with inputs frm the participants and thers in the cmmunity as apprpriate will refine the draft criteria. A final draft will be psted fr public cnsultatin. 12

14 e) Organizing a fllw up cnsultatin in the Fall f 2017 The prpsed bjective will be t finalize sme f the deliberatins nted abve and review the prgress with the agreed steps. Lcatin and date will be decided thrugh 13