Review

Development of against

Gregory A Poland, Richard B Kennedy, Inna G Ovsyannikova, Ricardo Palacios, Paulo Lee Ho, Jorge Kalil

Zika virus is an emerging pathogen of substantial public health concern to human beings. Although most are Lancet Infect Dis 2018; asymptomatic or present with benign, self-limited symptoms, a small percentage of patients have complications, such 18: e211–19 as congenital anomalies in the developing fetus of pregnant women infected with the virus and neurological Published Online complications (eg, Guillain-Barré syndrome). To date, there is no , antiviral drug, or other modality available to January 25, 2018 http://dx.doi.org/10.1016/ prevent or treat Zika virus . In this Review, we examine vaccine development efforts for Zika virus to date and S1473-3099(18)30063-X research gaps in the development of candidate vaccines against Zika virus. Top research priorities should include Mayo Clinic Vaccine Research development of a better understanding of immunity to Zika virus to establish clear correlates of protection; determination Group, Mayo Clinic, Rochester, of what effect, if any, Zika vaccine-induced immune responses will have on subsequent infection; MN , USA (Prof G A Poland MD, evaluation of vaccine immunogenicity and efficacy in healthy adults and in the various subpopulations affected by Zika Prof R B Kennedy PhD, Prof I G Ovsyannikova PhD); virus infection (children, pregnant women, women of childbearing age, and eldery people); and identification of the Instituto Butantan, São Paulo, molecular mechanisms that underlie birth defects and neurological sequelae related to Zika virus. Brazil (R Palacios MD, P L Ho PhD); and Medical School Introduction Review, we evaluate those efforts and outline scientific, and Heart Institute, University of São Paulo, São Paulo, Brazil Sometime in the early 1900s, probably in Uganda, the regulatory, and ethical issues related to the development (Prof J Kalil MD) ancestral Zika virus emerged and went undetected for of a vaccine against Zika virus. Correspondence to: decades. It was not until 1947, when sentinel rhesus Dr Gregory A Poland, Mayo monkey number 776 had a febrile illness, that researchers Zika virus Vaccine Research Group, looking for the yellow fever virus at the Ugandan Viral Zika virus is an arbovirus that belongs to the Flaviviridae Mayo Clinic, Rochester, 1 MN 55905, USA Research station identified Zika virus. Ugandans from family, which includes dengue, West Nile, and yellow fever [email protected] four different locations were studied and six (6·1%) of viruses. Zika virus has a single-stranded, positive sense 99 were positive for antibodies to the virus. At the time, RNA genome that encodes structural (capsid, precursor the importance of this emerging virus was unknown membrane [prM], and envelope [E]) and non-structural and unanticipated. The response to Zika virus as a proteins (1, 2A, 2B, 3, 4A, 4B, and 5). The genome is curiosity was justified, because at the time there were no translated as a single polyprotein and then is proteolytically known infections or outbreaks of importance in human cleaved into individual proteins. The prM, E, and non- beings. structural protein 1 have the greatest sequence divergence The first human case of Zika infection was reported in from other flaviviruses. Three main lineages of Zika virus 1952.2 Although Zika infection had been reported a decade have been found, east African, west African, and Asian. earlier,3,4 this infection was later identified as Spondweni These strains might be differentially associated with the virus.5 Other isolated cases of Zika infection were reported neurological defects seen during the outbreaks in the through the 1960s and 1970s.1,6 Nonetheless, Zika virus was Pacific and South America. Strains from Polynesia and not recognised as a human pathogen of importance until South America have high degrees of similarity (about 2007, when an outbreak of rash and fever occurred on Yap 99% sequence identity),9 while strains from the western Islands and led to 49 confirmed—and 59 probable—cases hemisphere are about 89% identical to the African of Zika virus infection in human beings.7 At the time, a genotypes.10 household survey of 557 people revealed that 412 (74%) had The virus was first isolated from rhesus macaques and antibodies for Zika virus. A larger second outbreak was subsequently found in mosquito vectors (including occurred in 2013–14 in French Polynesia, involving Aedes aegypti, Anopheles spp, and Mansonia spp).1 Specific 333 confirmed cases and 19 000–32 000 suspected cases.8 antibodies against Zika virus have been found in sheep, For the first time, Guillain-Barré syndrome was observed cows, goats, horses, water buffalo, ducks, rodents, bats, in conjunction with these infections. During the non-human primates, and other wild animals.11–17 This 2014–15 outbreaks of Zika virus in Brazil, public health large and diverse pool of animal hosts facilitates the authorities reported a noticeable increase in children born spread of the virus and can contribute to the emergence with microcephaly. of recombinant strains. Since 2007, local transmission has been documented in Barba-Spaeth and colleagues18 studied the crystal 52 countries and territories, including in continental structure of the Zika virus E protein complexed to USA. Despite the scale of the outbreak, it was not until two cross-reactive antibodies specific for dengue virus, February, 2016, that WHO declared Zika virus to be a which provided important data on conserved epitopes Public Health Emergency of International Concern. targeted by cross neutralising antibodies. Since then, the Before this announcement, there were no known efforts crystal structures of the mature19 and immature20 Zika to develop a vaccine against the Zika virus. Since then, virus have been resolved, allowing us to evaluate the researchers in this field have tried to understand Zika effect of mutations that have occurred since the start of virus and to develop a vaccine against Zika virus. In this the epidemic, to identify conserved regions with other www.thelancet.com/infection Vol 18 July 2018 e211 Review

flaviviruses, to map neutralising epitopes, and to better Zika virus is mostly suppressed. As a result, sexual understand viral attachment and entry. transmission of the virus has been documented,33,34 and perinatal transmission has been described during Zika virus infection pregnancy.35,36 Thus, vaccines and treatments designed to Zika virus, which has rapidly emerged as a threat to fight Zika virus infection in all organs is an important public health, is transmitted by A aegypti mosquitoes.21 consideration in vaccine development. An improved Zika virus spread beyond its historical range in equatorial understanding of interactions between the virus and host Africa and Asia to the Federated States of Micronesia will inform the public health response to this pathogen. (Yap Islands). In 2015, the virus spread to the Caribbean, South America, and Central America, where it has Immune responses to Zika virus caused epidemics.22 22·7 million people in the USA live Many cellular proteins could act as potential entry receptors in areas with active mosquito populations that are (dendritic-cell-specific ICAM-3 grabbing non-integrin, conducive to year-round transmission of Zika virus.23 The Axl receptor tyrosine kinase, hepatitis A virus cellular Aedes species of mosquitoes known to transmit Zika receptor 1, tyrosine protein kinase 3).37,38 Cellular infection virus—A aegypti and A albopictus—are present in more with Zika virus is characterised by expression of Toll-like than 30 states and territories in the USA. Furthermore, receptor 3, retinoic acid inducible gene-I, and melanoma 60% of people in the USA live in areas that might allow differentiation associated gene 5, secretion of type 1 seasonal transmission of Zika virus due to the and type 2 interferons, and upregulation of interferon- geographical distribution of Aedes spp mosquitos.23 The stimulated genes that mediate antiviral responses (OAS2, mosquito vectors carrying Zika virus are expanding in ISG15, or MX1).38 Glycosaminoglycans can aid the geographical location, and currently cover Africa, South attachment and entry of Zika virus.39 and Central America, India, Singapore, Puerto Rico, and Initial reports indicated that immune responses to the USA, and will probably spread to other parts of Asia.24 Zika virus were similar to that of other flaviviruses in that As of March 10, 2017, people in 84 countries and neutralising antibodies (primarily directed at E protein) territories had been infected with Zika virus.25 are thought to play a major part in protection against 80% of Zika virus infections are asymptomatic, while infection.40–43 In fact, there is considerable serological the other 20% of infections typically present as a mild crossreactivity between Zika virus and other flaviviruses, form of dengue-like disease with mild fever, skin rash, and this reactivity has hindered the use of diagnostic muscle and joint pain, conjunctivitis, vomiting, tests and seroprevalence studies in geographical areas headache, malaise, prostration, and oedema of the hands where antibody responses to other flaviviruses are also and feet.26 These symptoms usually last for 2–7 days, and present.1,44 In some geographical areas, monoinfection the incubation period of Zika virus typically ranges from with Zika virus is less common than co-infection with 3 days to 12 days. other flaviviruses.45 Zika virus infection has been associated with serious The E protein of flaviviruses (including Zika virus) complications during pregnancy. In pregnant women, contains three domains, with most neutralising Zika virus infection can cause severe fetal neuro­ antibodies targeting determinants in domain 3 or the malformations or fetal death.27 The increase in Brazilian fusogenic loop of domain 2.46 Multiple monoclonal and French Polynesian epidemics of congenital micro­ antibodies targeting the Zika virus E protein have been cephaly with related malformations28 was caused, at least developed; these antibodies provide varying amounts of in part, by in-utero infection of Zika virus. Zika virus protection after passive transfer in lethal mouse models infection has also been associated with an increased of infection.46 These results show the important role that incidence of Guillain-Barré syndrome, which is an humoral responses have in protective immunity. autoimmune neurological complication characterised by A single dose of purified inactivated virus vaccine, or a muscle weakness, motor dysfunction and, in some cases, plasmid DNA vaccine encoding the prM and E proteins as paralysis.29,30 immunogens, showed complete protection against Zika The Zika viral genome has been found in bodily fluids virus challenge in mouse and rhesus monkey models of (eg, semen, urine, breast milk, saliva, and amniotic fluid) infection.47 In addition, a rhesus adenovirus (serotype 52) for weeks after infection. In particular, Zika virus RNA vaccine led to complete protection after one dose in rhesus can be detected in semen for at least 62 days once an monkeys.40 This DNA vaccine also elicited protective individual is infected, and Zika virus RNA can be found immunity in animals that were depleted of CD4-positive in semen when it is undetectable in blood.31 For example, or CD8-positive T cells before viral challenge, indicating of the 23 men infected with the virus in the UK, high that humoral immunity is sufficient to mediate clinical concentrations of Zika virus RNA were detected in the protection. These experiments have begun to identify a semen of 13 (57%) men.32 The persistence of Zika virus in possible serological correlate of protection in E antibody semen suggests possible risks for sexual transmission. concentrations. It is still unknown whether T cell help is Although the testes have some T cells and macrophages, necessary during the initiation of B cell responses that are their ability to develop an immune response against specific to Zika virus. e212 www.thelancet.com/infection Vol 18 July 2018 Review

To date, it is unknown if Zika virus antibodies will showed how Zika virus adapted to mice after high show antibody-dependent enhancement of disease as lethality rates in early passages.55 Later, pathogenicity in seen in dengue virus (where prior infection with one the mouse model was better described,56 and other serotype leads to the development of an antibody non-human primates were reported as susceptible.13,15,17 response that does not fully neutralise the second Susceptibility and pathogenicity of a virus in animal serotype, but instead facilitates viral infection of species is a result of host–virus interaction. This macrophages, which leads to more severe disease), how relationship is dynamic, even within the same species, as immune responses developed against Zika virus will demonstrated by the increasing pathogenicity of Zika affect subsequent infection with dengue virus (or vice virus in human beings. The use of animal models for versa), or if the neurological abnormalities, such as Zika virus infection and vaccine-induced immune Guillain-Barré syndrome observed in some patients, are response and efficacy research contributes greatly to a result of direct viral infection or a side-effect of the vaccine development. immune response. The first model of successful intrauterine infection Another issue that could affect the immunity caused by with a Brazilian isolate of Zika virus was in SJL mice at against Zika virus is original antigenic sin, day 10–13 of gestation.57 Offspring born after challenge which occurs when the immune response to a second had intrauterine growth restriction and severe Zika virus infection is dominated by the immunological memory infection of the brain. Notably, C57BL/6 pups were response to epitopes seen during a previous infection. unaffected by a similar challenge during pregnancy. For The immune system targets the antigens seen during the both mice strains, mothers who were exposed to the initial infection, but does so at the expense of an immune virus did not develop any relevant sign of infection.57 In response to new antigens presented during the second another model, injection of Zika virus into the brain of infection. Although this effect is typically seen in C57 embryos caused substantial brain damage and influenza infection, it has also been observed with intrauterine infection.58 dengue.48 The extent to which pre-existing immunity to Knockout mice with deficiencies for immune receptors other flaviviruses influences the response to Zika virus are often used to create animal models; however, infection is unclear and needs further investigation. such host modifications might trigger mechanisms Studies on the effect of original antigenic sin on that misrepresent viral pathological features. In one vaccination have conflicting results.49,50 Given the potential experiment, mice from the same origin (129Sv/Ev strain effect this event could have on vaccine efficacy in areas of and its mutated congenic A129 strain) that were missing the world where multiple flaviviruses are endemic and co- type 1 interferon receptors received 1 × 10⁶ plaque forming infections are common, vaccine formulations need to be units of Zika virus subcutaneously. Infection was lethal designed to consider this effect.51 for mutated mice, but mice with the non-mutated strain Given the extensive antibody cross-reactivity that has had no symptoms.59 In another parallel experiment, been seen between flaviviruses, it is important to consider AG129 mice lacking type 1 and type 2 interferon receptors how broadly protective immune responses to one strain had 100% lethal Zika virus infection after intraperitoneal of Zika virus are against other strains. Zika virus has been and footpad injection of doses ranging from 1 to classified into three genotypes, west African, east African, 1 × 10⁵ plaque forming units.60 The main necropsy and Asian; the E protein aminoacid sequence is conserved finding in both experiments was meningoencephalitis, at more than 99% across all three lineages.10 In-vitro particularly in the hippocampus. Another experiment studies and several challenge studies40,43,46,52 have shown infected both knockout mice (A129 and AG129), but had that extensive cross-reactivity exists. Osuna and analogous results and no substantial differences between colleagues53 showed that infection with a Zika virus isolate both strains.61 Those experiments highlighted the from Thailand protected against subsequent challenge importance of interferon in host responses to Zika virus. with an isolate from Puerto Rico. Both of those strains are Tests in mice with low production of interferon α and members of the Asian lineage. Aliota and colleagues54 interferon β confirmed this finding.62 showed that immunisation with the prototype MR766 With regard to the initial isolation in Uganda, the strain from Uganda provides protective immunity against rhesus monkey model was susceptible to the Asian a heterologous French Polynesian strain, Zika virus-FP. lineage of Zika virus. The virus was found in bodily Notably, this finding showed that antibodies elicited fluids, similar to human clinical findings. Subsequent during infection with an east African lineage of Zika virus exposure to the virus on day 70 after initial infection cross protect against Zika virus from the Asian lineage. showed immune protection. Notably, two pregnant monkeys infected during the first trimester had Animal models for vaccine development prolonged viraemia, but amniocentesis samples were The first isolation of Zika virus in 1947 did not occur in negative for Zika virus.63 An update on the Zika For more on the Zika natural hosts, but in animal models. Swiss albino mice Open-Research Portal of one birthed monkey showed Open-Research Portal see https://zika.labkey.com were intracerebrally inoculated with sera from non- that Zika virus was found only in the fetus’ optical nerve, human primates (rhesus monkeys). This early paper axillary lymph node, and bone marrow. A preliminary www.thelancet.com/infection Vol 18 July 2018 e213 Review

report found Zika virus in a wild capuchin monkey and a and antiviral drug therapy.65 Nevertheless, impaired marmoset from northeast Brazil.64 This finding not only immune responses can impair vaccine development. identifies putative reservoirs, but also potential animal For active immunisation, susceptible models with models. Although non-human primates are the most preserved immune responses are preferable. Larocca and similar to human beings in terms of virus–host colleagues47 showed that the BALB/c mouse is analogous interaction, and for the evaluation of safety and efficacy to the SJL mouse as a model for Zika virus infection and of vaccines and therapies, there are limitations to their could mimic the natural mild infection in human beings, use, such as animal availability, cost, and ethical issues. with a small window of viraemia detected in the blood. Mouse models with substantially compromised Both mice strains received two types of Zika virus immune responses (ie, interferon activity) are useful vaccines (prM and E DNA and purified inactivated).47 for the evaluation of the effects of passive immunity Yi and colleagues66 found that this same vaccine elicits sterile immunity in mice reconstituted with human Status Type of vaccine haemopoietic stem cells; use of these mice should be further evaluated to determine their usefulness as animal Inovio In phase 1 clinical trials DNA vaccine models. National Institutes of In phase 1 clinical trials DNA vaccine; live vesicular stomatitis virus Health recombinant (early R&D); live attenuated Zika virus (early R&D) Vaccine efforts Walter Reed Army In phase 1 clinical trials Whole, purified, inactivated virus Given the efficacy and success of the , Institute of Research and the recent successful development of a vaccine and Sanofi Pasteur against dengue virus, the development of a safe and Butantan Institute In phase 1 clinical trials; Live, dengue virus-vectored vaccine expressing effective vaccine against Zika virus is a realistic early stage research precursor membrane and envelope proteins; purified inactivated virus possibility, and multiple vaccine candidates are in various 67 Bharat Preclinical animal studies Purified inactivated virus; virus-like particle expressing stages of development. A WHO report from March, polyprotein 2016, showed that there were 18 known Zika virus NewLink Genetics Preclinical animal studies Purified inactivated virus vaccines in development, although none had progressed PaxVax Preclinical animal studies Purified inactivated virus beyond early preclinical development at that point.68 Novavax Preclinical animal studies Protein nanoparticle vaccine These vaccine candidates include various technologies Replikins Preclinical animal studies Synthetic and approaches, including inactivated Zika virus, Pharos Biologicals Preclinical animal studies DNA vaccine attenuated Zika virus strains, live or inactivated viral Bio-Manguinhos Early stage research Purified inactivated virus; yellow fever 17DD chimera; recombinants expressing Zika virus proteins (eg, dengue virus-like particle; DNA virus, modified vaccinia virus Ankara, adenovirus, US Centers for Disease Early stage research Virus-like particle expressing Zika virus DNA; live lentivirus, measles virus), viral-like particles expressing Control and Prevention adenovirus recombinant Zika virus membrane proteins, recombinant protein CureVac Early stage research Thermostable mRNA-based vaccine vaccines, DNA plasmid vaccines, mRNA-based vaccines, Geovax Early stage research Live modified vaccinia ankara recombinant protein-nanoparticle conjugates, and peptide-based GlaxoSmithKline Early stage research Self-amplifying mRNA platform; whole, inactivated virus vaccines.69 These products are being produced and tested Hawaii Biotech Early stage research Alhydrogel and recombinant protein by pharmaceutical and vaccine companies, universities, Oxford University Early stage research Live adenovirus recombinant government agencies, private foundations, and multi- Protein Sciences Early stage research Recombinant envelope protein institute collaborations (table). In addition, several large Sanofi Early stage research Yellow fever 17D chimera vaccine companies are evaluating whether their existing Sementis Early stage research Live poxvirus recombinant vaccine platforms can be adapted to include protection Themis Bioscience Early stage research Live measles recombinant against Zika virus. Valneva Early stage research Purified inactivated virus Mayo Clinic Vaccine Early stage research Naturally processed and HLA-presented Zika virus Whole virus Research Group peptides packaged with biodegradable nanoparticles A whole virus, formalin-inactivated vaccine (named Early stage research Lipid nanoparticle-delivered mRNA ZPIV), developed by the Walter Reed Army Institute of Emergent Biosolutions Early stage research Inactivated, whole virus Research (Silver Spring, MD, USA), has shown protective Institut Pasteur of Early stage research Recombinant subunit virus-like particle efficacy in mice after a single injection (with or without Shanghai aluminium).47 This vaccine caused increased antibody Takeda Early stage research Alum adjuvanted, inactivated whole virus titres and superior protection when administered by the Jenner Institute Early stage research Simian adenovirus vector intramuscular route compared with the subcutaneous VBI Vaccines Early stage research Virus-like particle containing envelope and non- route. This same vaccine was also tested in non-human structural 1 proteins primates.40 Rhesus macaques immunised with the Vaxart Early stage research Recombinant oral vaccine inactivated vaccine developed humoral and cellular R&D=research and development. immunity, with no detectable virus present in blood, urine, CSF, or other body fluids after exposure to live Table: Zika vaccines in development Zika virus. High doses of purified IgG from vaccinated e214 www.thelancet.com/infection Vol 18 July 2018 Review

animals, which was adoptively transferred into naive exposed to the virus. Another study in C57BL/6 mice animals, provided complete protection upon challenge. tested two additional vaccines (a recombinant subunit ZPIV is currently undergoing phase 1 clinical testing in vaccine and an adenovirus vectored vaccine), both human beings with no history of flavivirus infection.40 containing the Zika virus E protein fused to the T4 Follow-up clinical trials to evaluate antigen dose, number fibritin foldon trimerisation domain. Although both of , and efficacy in individuals exposed to vaccines elicited neutralising antibodies, the titres in dengue virus are ongoing. mice that received the were four to Additional inactivated vaccine formulations by 16 times lower than in mice given the adenovirus- GlaxoSmithKline, PaxVax, NewLink Genetics, and Bharat vectored vaccine.76 In this study, pups born to mice Biotech are in preclinical development.70 immunised with the adenovirus vector vaccine were protected against lethal intraperitoneal exposure to the Nucleic acid vaccines virus. These studies provide key data on the immune A plasmid DNA vaccine has also shown protection in response mechanisms needed for protection and help immunised mice.47 The plasmid vaccine encodes the prM preparations for human testing of these vaccine and E antigens from the BeH815744 strain of Zika virus formulations. Additional vectored vaccines that use and was given as a single dose intramuscularly. adenovirus, vesicular stomatitis virus, measles, and other Vaccinated mice had high titres of neutralising antibody, backbones are in the early stages of research and detectable, virus-specific CD4-positive and CD8-positive development. The US National Institutes of Health and T-cell responses, and showed protection (lack of the Butantan Institute have started a phase 1 detectable viraemia) against live virus. Notably, a plasmid of a live against Zika virus; this variant lacking the prM antigen was less antigenic, vaccine is based on the tetravalent that is eliciting significantly lower neutralising antibody titres undergoing clinical trials in Brazil.70 than did the plasmid with the prM antigen. A subsequent study of rhesus macaques had similar results.40 Rhesus Additional vaccines macaques immunised with the inactivated vaccine Protein-based and peptide-based vaccines are also being developed humoral and cellular immunity, with no investigated by many research groups. The Mayo Clinic detectable virus present in blood, urine, CSF, or other Vaccine Research Group is conducting initial studies to bodily fluids following exposure to live Zika virus. High identify HLA class I and II epitopes from Zika virus to doses of purified IgG from vaccinated animals, which evaluate peptide-based vaccines. This research is being was adoptively transferred into naive animals, provided done in collaboration with Iowa State University and the complete protection upon exposure to the virus. A NanoVaccine Initiative. The aim is to develop an prime-boost vaccination regimen (4 weeks apart) with immunogenic, nanoparticle-encapsulated peptide-based 5 mg of the prM and E DNA vaccine was necessary for vaccine that is easily manufactured, stable, requires no optimal development of neutralising antibody responses cold chain, and is safe for use in pregnant women and and provided complete protection against viral challenge. immunocompromised individuals. This DNA vaccine is also undergoing clinical trials.71 Given the gaps in knowledge related to Zika virus Pardi and colleagues72 have reported that a single dose infection and immunity, and the public health concerns (50 µg) of a nucleoside-modified RNA vaccine, which related to pregnancy and fetal malformations, the ideal expresses the prM and E proteins, provides protection in vaccine will be based on a non-replicating platform that both mice and non-human primates. The resulting is safe for use during pregnancy. These platforms include immune responses also protected against heterologous inactivated whole virus, subunit, or mRNA-based or strains of Zika virus. DNA-based vaccines that express selected viral proteins. In July, 2016, Inovio Pharmaceuticals started a If there is adequate cross-strain neutralisation or phase 1 clinical trial of its DNA-based vaccine protection, then the use of strains of Zika virus that are (GLS-5700).73 The US National Institute of Allergy and not associated with fetal malformations (eg, those from Infectious Diseases has also developed an investigational the African lineage) would be better from a safety DNA vaccine that is undergoing phase 1 testing in standpoint. Populations with distinctly different risks human beings. This vaccine candidate is based on a exist, including pregnant women versus healthy men, vaccine for .74 A follow-up phase 2 clinical infants and children (immature immune system) versus trial in Puerto Rico, where Zika virus is endemic, began women of childbearing age, healthy adults versus the in August, 2016.75 elderly (immunosenescence and increased risk of Guillain-Barré syndrome), people who have not been Vectored vaccines exposed to Zika virus (or other flaviviridae) versus those In a non-human primate study, a rhesus adenovirus with a history of infection—not to mention the grow­ing serotype 52 vectored-vaccine (RhAd52-prM-E) was also population worldwide of immunocompromised evaluated.72 A single dose of the vaccine caused a robust individuals. Because of this wide range of requirements, antibody response with 100% protection in animals multiple vaccine formulations could be necessary. Much www.thelancet.com/infection Vol 18 July 2018 e215 Review

like the current situation with influenza vaccines, From a regulatory perspective, a vaccine might be vaccines against Zika virus might be geared toward, and more straightforward to license for healthy children and optimised for, specific cohorts of the population. adults with a strategy aimed at mass vaccination before From an efficacy standpoint, the optimal endpoints childbearing years. Nonetheless, because of the need to be determined. Possible endpoints include a substantial morbidity associated with Zika virus infection notable reduction in clinical disease, the prevention of in pregnancy, pregnant women and women of viraemia in vaccine recipients, or prevention of either childbearing age should be a top priority for vaccine neurological complications or congenital Zika virus development and evaluation. Critical barriers to this syndrome. To date, the vaccine candidates in clinical prioritisation have been highlighted by Omer and Beigi,77 trials have shown efficacy endpoints of prevention of including a broadly accepted ethical framework for infection in animal models. Secondary endpoints tested vaccine research in this group of individuals, a knowledge in animal models include putative immunological gap in understanding immune responses to vaccination correlates of protection (ie, neutralising antibody titres in early pregnancy, regulatory issues linked to labelling above a certain threshold). Confirmation of these of vaccines for use in pregnancy, an absence of standard correlates as actual correlates of protection in large definitions of outcomes in regard to evaluation of randomised clinical trials will facilitate vaccine vaccines in pregnancy,78 difficulties in the use of real-time development and efficacy testing. In the absence of safety assessments of risk versus benefit, and poor data specific data on Zika virus, immunological correlates in regard to the use of baseline rates of outcomes— from other flaviviruses might also be a useful starting particularly in areas with other co-infections associated point, especially for the initial development of animal with adverse birth outcomes and in areas with inadequate models. However, there is no consensus on correlates of public health resources to capture such data.77 protection with other flaviviruses—for example, there are Live attenuated vaccine designs raise serious issues no generally accepted correlates of protection in dengue. when administered to potentially pregnant women or Such correlates are complicated by the need for cellular in immunocompromised individuals, particularly in immunity and the need to define what the specified geographical areas where inexpensive and rapid testing endpoint of protection is—clinical disease, subclinical for early pregnancy might not be readily available. The disease or infection, or congenital Zika virus syndrome. use of passive protection for unborn and newborn Finally, measures of vaccine efficacy should consider children by maternal immunisation and subsequent differences in mode of infection, such as mosquito-borne neonatal or infant protection might be needed in some transmitted disease, sexually transmitted (mucosal geographical settings. exposure) infection, and maternally transmitted infection. Different modes of infection might require Research agenda for Zika virus vaccination different types of vaccine solutions and different clinical We have previously published a brief review of vaccine endpoints or measures of efficacy. For example, response, development, and regulatory issues that are subclinical Zika virus infection could still lead to useful for the development of Zika vaccines (panel).79 congenital Zika virus syndrome, thus requiring study There are also other considerations.80 For example, design endpoints that take such considerations into low-cost vaccines that are suitable for wide-scale use in effect, including sub­clinical endpoints of Zika virus less-developed countries are key to an effective global (seroconversion, frequent serological testing, and response. Other issues that require attention from a focused use of diagnostic tests capable of detecting early public health point of view include ease of vaccine and subclinical viraemia). administration, potential need for a cold chain, multiple doses for protective immunity, duration of immune Regulatory hurdles response over time, identification of possible correlates Although regulatory pathways for certain types of of protection, potential antibody-dependent enhancement vaccines are clear (inactivated, live attenuated, subunit, of disease due to immunisation of people in areas where and recombinant vaccines), development and regulatory flaviviruses are endemic, and classification of vaccine- pathways for other types of vaccine (DNA and peptide- induced immunity and wild virus-induced immunity. based vaccines) are less clear. Vaccine-induced adverse Another area of interest is whether vaccines for Zika effects in pregnant women and the ability to safely virus—based on type of vaccine, immunogenicity, and develop protective immune responses among important immune durability and persistence—can lead to subgroups of the population are important issues to interrupted epidemic transmission or sustained herd consider. Such subgroups include infants, children, immunity, or both, in the population. Research is needed pregnant women, immunocompromised individuals, into the immune response across the population in people with pre-existing antibody to other closely related terms of candidate vaccines for Zika virus. There are no arboviruses, population groups in less-developed data on what the immune response and outcome might countries with co-infections and malnutrition, and be if a pregnant woman or man with detectable Zika immunosenescent individuals. virus in the testes were immunised, or on other e216 www.thelancet.com/infection Vol 18 July 2018 Review

reproductive toxicology endpoints. The growing population of immunosenescent individuals and Panel: Key challenges for Zika virus vaccination immunocompromised people needs to be considered • Antibody-mediated immune enhancement of dengue during vaccine development. How immunity to candidate virus infection needs to be avoided vaccines might interact with pre-existing flavivirus • The vaccine should be able to elicit protective immunity, immunity is key to the safety and efficacy of these regardless of previous exposure to dengue virus vaccines. Substantial concern exists over the ethical use • The vaccine needs to be safe for vulnerable populations, of challenge trials to determine vaccine efficacy and including pre-pubescent children, pregnant women, potentially safety. Finally, once these vaccines are and men and women of childbearing age developed, economic and public health policies should • Protective immunity should be transferred to the be aligned with the properties of the vaccines to maximise developing fetus and newborn child the benefits of such vaccines. In this regard, WHO has • Vaccination should not cause neurological side-effects, outlined an initial Target Product Profile for vaccines especially given the link between Zika virus and against Zika virus—particularly for use during public Guillain-Barré syndrome 81 health emergencies. • To date, there is not an established correlate of protection • The vaccine should protect healthy adults, young children, Conclusions pregnant women, and unborn fetuses—who could all Affordable, easily available, and highly effective antiviral require a different level of immunity drugs and vaccines that prevent Zika virus infection are • Clinical efficacy is very difficult to assesswhen 80% of key to protecting the public, particularly women of infections are asymptomatic childbearing age. Even the potential availability of antiviral drugs would not prevent asymptomatic infection leading to fetal abnormalities in pregnant women, or Search strategy and selection criteria other complications such as Guillain-Barré syndrome. Regardless, it is not feasible to provide continuous We searched PubMed using the following terms “ZIKV“, “Zika“, chemoprophylaxis to entire populations. For this reason, OR “Zika virus“ alone and in combination with “vaccine“, safe and effective vaccines against Zika virus with “vaccine development“, “infection“, “pathogenesis“, long-term, if not lifelong, immunity need to be developed. “outbreak“, “pregnancy“, “birth defect“, OR “animal model“. To be useful, such vaccines need to be inexpensive We searched from inception to Nov 6, 2017. All types of articles enough to be widely used, particularly in low-resource were included. We only considered articles published in English. countries where most outbreaks occur. We also searched clinical trial websites (ie, ClinicalTrials.gov), Of concern is the possibility of the Zika virus mutating PubMed, and Google Scholar for press releases about Zika virus in a way that increases its pathogenicity, especially products. among pregnant women or young children, elderly people, and immunocompromised people. This situation adverse events, such as Guillain-Barré syndrome after could require different types of vaccine for different wild virus infection. It is unknown who will cover these subpopulations. For example, among healthy, non- costs. Clinical trials that are designed to show efficacy pregnant individuals, a live attenuated viral approach against infection can be somewhat straightforward, but might be ideal; whereas a subunit vaccine might be the use of efficacy endpoints against complications—such the best option for pregnant women. Among elderly as congenital Zika virus syndrome, other congenital and immunocompromised individuals, an adjuvanted anomalies, and Guillain-Barré syndrome—could require vaccine might be most immunogenic and effective. Such extremely large numbers of study volunteers to have issues need to be considered as the scientific community sufficient statistical power. Studies designed to determine learn more about the generation of protective immunity possible congenital adverse effects are also complex and to this virus, and the ways in which it could change will probably require very large numbers of participants to during outbreaks in naive populations or in response to reduce concerns about the safety of the therapy. further selection pressure for mutation. At least one suggestion for funding such costs has been An ongoing issue is the lack of research funding for the call to develop a global vaccine development fund.82 vaccine development. The US Government has approved The benefit of such a design is that the vaccines developed US$1·1 billion in funding for Zika virus. WHO, Google, as a result are funded by all the countries that stand to and The Bill & Melinda Gates Foundation have also benefit from having a safe and effective vaccine. pledged financial support. If Zika virus infections, media Nonetheless, questions and concerns remain regarding attention, and public health pressure reduce, efforts to who owns the intellectual property of such research and develop a vaccine could decline, as happened with vaccines who gains financially from marketing a much needed for Ebola virus. Expensive clinical trials will be necessary vaccine. because of issues with immunising pregnant women, We conclude that vaccines for Zika virus are urgently young children, the elderly, and concerns over possible needed to protect the health of the public. Funding that www.thelancet.com/infection Vol 18 July 2018 e217 Review

enables rapid vaccine development and phase 1–3 clinical 16 Darwish MA, Hoogstraal H, Roberts TJ, Ahmed IP, Omar F. trials need to made available. Notably, there are an A sero-epidemiological survey for certain arboviruses (Togaviridae) in Pakistan. Trans R Soc Trop Med Hyg 1983; 77: 442–45. increasing number of outbreaks of infectious diseases 17 Wolfe ND, Kilbourn AM, Karesh WB, et al. Sylvatic transmission of that threaten public health on a global scale. The global arboviruses among Bornean orangutans. Am J Trop Med Hyg 2001; health community continue to handle these outbreaks as 64: 310–16. 18 Barba-Spaeth G, Dejnirattisai W, Rouvinski A, et al. Structural basis exceptions, failing to recognise the need for a global and of potent Zika-dengue virus antibody cross-neutralization. NW16; efficiently coordinated response strategy. Without such 536: 48–53. efforts, the eventual development of a vaccine against 19 Sirohi D, Chen Z, Sun L, et al. The 3.8 A resolution cryo-EM Zika virus will be recognised as too little too late. structure of Zika virus. Science 2016; 352: 467–70. 20 Prasad VM, Miller AS, Klose T, et al. Structure of the immature Contributors Zika virus at 9 A resolution. Nat Struct Mol Biol 2017; 24: 184–86. GAP, RBK, IGO, RP, PLH, and JK drafted sections of the Review 21 Ioos S, Mallet HP, Leparc Goffart I, authierG V, Cardoso T, according to their own areas of expertise and their literature searches. Herida M. Current Zika virus and recent epidemics. Med Maladies Infect 2014; 44: 302–07. Declaration of interests 22 Zanluca C, Melo VC, Mosimann AL, Santos GI, Santos CN, Luz K. GAP is the chair of a Safety Evaluation Committee for novel First report of autochthonous transmission of Zika virus in Brazil. investigational vaccine trials being conducted by Merck Research Mem Inst Oswaldo Cruz 2015; 110: 569–72. Laboratories and consultants on vaccine development for Merck & Co, 23 Bogoch II, Brady OJ, Kraemer MU, et al. Anticipating the international Avianax, Dynavax, Novartis Vaccines and Therapeutics, Adjuvance spread of Zika virus from Brazil. Lancet 2016; 387: 335–36. Technologies, Seqirus, and Protein Sciences. GAP and IGO have 24 Gulich GA. Epidemiology, driving factors, transmission and control three patents related to vaccinia and measles peptide research. RBK has options of Zika virus: a review. J Infect Dis Ther 2016; 4: 278. received funding from Merck Research Laboratories to study waning 25 WHO. Zika situation report. Zika virus, microcephaly and immunity to . All other authors declare no competing Guillain-Barre syndrome. 2017. http://www.who.int/emergencies/ interests. zika-virus/situation-report/10-march-2017/en/ Acknowledgments (accessed Jan 15, 2018). We thank Caroline L Vitse (Mayo Clinic Vaccine Research Group) for 26 Shuaib W, Stanazai H, Abazid AG, Mattar AA. The reemergence of editorial assistance. Zika virus: a review on pathogenesis, clinical manifestations, diagnosis, and treatment. Am J Med 2016; 129: 879. References 27 Panchaud A, Stojanov M, Ammerdorffer A, Vouga M, Baud D. 1 Dick GW, Kitchen SF, Haddow AJ. Zika virus. I. Isolations and Emerging role of Zika virus in adverse fetal and neonatal outcomes. serological specificity. Trans R Soc Trop Med Hyg 1952; 46: 509–20. Clin Microbiol Rev 2016; 29: 659–94. 2 Kindhauser MK, Allen T, Frank V, Santhana RS, Dye C. Zika: the 28 White MK, Wollebo HS, David Beckham J, Tyler KL, Khalili K. origin and spread of a mosquito-borne virus. Bull World Health Zika virus: an emergent neuropathological agent. Ann Neurol 2016; Organ 2016; 94: 675–86C. 80: 479–89. 3 MacNamara FN. Zika virus: a report on three cases of human 29 Roze B, Najioullah F, Ferge JL, et al. Zika virus detection in urine infection during an epidemic of jaundic in Nigeria. from patients with Guillain-Barré syndrome on Martinique, Trans R Soc Trop Med Hyg 1954; 48: 139–45. January 2016. Euro Surveill 2016; 21: 30154. 4 Bearcroft WG. Zika virus infection experimentally induced in a 30 Cao-Lormeau VM, Blake A, Mons S, et al. Guillain-Barré syndrome human volunteer. Trans R Soc Trop Med Hyg 1956; 50: 442–48. outbreak associated with Zika virus infection in French Polynesia: 5 Simpson DI. Zika virus infection in man. Trans R Soc Trop Med Hyg a case-control study. Lancet 2016; 387: 1531–39. 1964; 58: 335–38. 31 Atkinson B, Hearn P, Afrough B, et al. Detection of Zika virus in 6 Filipe AR, Martins CM, Rocha H. Laboratory infection with Zika semen. Emerg Infect Dis 2016; 22: 940. virus after vaccination against yellow fever. Arch Gesamte Virusforsch 32 Atkinson B, Thorburn F, Petridou C, et al. Presence and persistence 1973; 43: 315–19. of Zika virus RNA in semen, United Kingdom, 2016. 7 Duffy MR, Chen TH, Hancock WT, et al. Zika virus outbreak on Emerg Infect Dis 2017; 23: 611–15. Yap Island, Federated States of Micronesia. N Engl J Med 2009; 33 Johansson MA, Mier YT-RL, Reefhuis J, Gilboa SM, Hills SL. 360: 2536–43. Zika and the risk of microcephaly. N Engl J Med 2016; 375: 1–4. 8 Roth A, Mercier A, Lepers C, et al. Concurrent outbreaks of dengue, 34 Hills SL, Russell K, Hennessey M, et al. Transmission of Zika virus and Zika virus infections—an unprecedented through sexual contact with travelers to areas of ongoing epidemic wave of mosquito-borne viruses in the Pacific 2012–2014. transmission—continental United States, 2016. Euro Surveill 2014; 19: 20929. MMWR Morb Mortal Wkly Rep 2016; 65: 215–16. 9 Gatherer D, Kohl A. Zika virus: a previously slow pandemic spreads 35 Oliveira Melo AS, Malinger G, Ximenes R, Szejnfeld PO, rapidly through the Americas. J Gen Virol 2016; 97: 269–73. Alves Sampaio S, Bispo de Filippis AM. Zika virus intrauterine 10 Lanciotti RS, Lambert AJ, Holodniy M, Saavedra S, infection causes fetal brain abnormality and microcephaly: tip of del Carmen Castillo Signor L. Phylogeny of Zika virus in Western the iceberg? Ultrasound Obstet Gynecol 2016; 47: 6–7. hemisphere, 2015. Emerg Infect Dis 2016; 22: 933–35. 36 Calvet G, Aguiar RS, Melo AS, et al. Detection and sequencing of 11 Vorou R. Zika virus, vectors, reservoirs, amplifying hosts, and their Zika virus from amniotic fluid of fetuses with microcephaly in potential to spread worldwide: what we know and what we should Brazil: a case study. Lancet Infect Dis 2016; 16: 653–60. investigate urgently. Int J Infect Dis 2016; 48: 85–90. 37 Meertens L, Labeau A, Dejarnac O, et al. Axl mediates ZIKA virus 12 Bres P. Recent data from serological surveys on the prevalence of entry in human glial cells and modulates innate immune arbovirus infections in Africa, with special reference to yellow fever. responses. Cell Rep 2017; 18: 324–33. Bull World Health Organ 1970; 43: 223–67. 38 Hamel R, Dejarnac O, Wichit S, et al. Biology of Zika virus infection 13 McCrae AW, Kirya BG. Yellow fever and Zika virus epizootics and in human skin cells. J Virol 2015; 89: 8880–96. enzootics in Uganda. Trans R Soc Trop Med Hyg 1982; 76: 552–62. 39 Kim SY, Zhao J, Liu X, et al. Interaction of Zika virus envelope 14 Olson JG, Ksiazek TG, Gubler DJ, et al. A survey for arboviral protein with glycosaminoglycans. Biochemistry 2017; 56: 1151–62. antibodies in sera of humans and animals in Lombok, Republic of 40 Abbink P, Larocca RA, De La Barrera RA, et al. Protective efficacy of Indonesia. Ann Trop Med Parasitol 1983; 77: 131–37. multiple vaccine platforms against Zika virus challenge in rhesus 15 Saluzzo JF, Ivanoff B, Languillat G, Georges AJ. Serological survey monkeys. Science 2016; 353: 1129–32. for arbovirus antibodies in the human and simian populations of 41 Dai L, Song J, Lu X, et al. Structures of the Zika virus envelope the South-East of Gabon. Bull Soc Pathol Exot Filiales 1982; protein and its complex with a flavivirus broadly protective antibody. 75: 262–66. Cell Host Microbe 2016; 19: 696–704.

e218 www.thelancet.com/infection Vol 18 July 2018 Review

42 Dowd KA, Ko SY, Morabito KM, et al. Rapid development of a DNA 64 Favoretto S, Araujo D, Oliveira DS, et al. First detection of Zika vaccine for Zika virus. Science 2016; 354: 237–40. virus in neotropical primates in Brazil: a possible new reservoir. 43 Sapparapu G, Fernandez E, Kose N, et al. Neutralizing human bioRxiv 2016; 049395. antibodies prevent Zika virus replication and fetal disease in mice. 65 Zmurko J, Marques RE, Schols D, Verbeken E, Kaptein SJ, Neyts J. Nature 2016; 540: 443–47. The viral polymerase inhibitor 7-deaza-2’-C-methyladenosine is a 44 Priyamvada L, Quicke KM, Hudson WH, et al. Human antibody potent Inhibitor of in vitro Zika virus replication and delays disease responses after dengue virus infection are highly cross-reactive to progression in a robust mouse infection model. PLoS Negl Trop Dis Zika virus. Proc Natl Acad Sci USA 2016; 113: 7852–57. 2016; 10: e0004695. 45 Rodriguez-Morales AJ, Villamil-Gomez WE, Franco-Paredes C. 66 Yi G, Xu X, Abraham S, et al. A DNA vaccine protects human The arboviral burden of disease caused by co-circulation and immune cells against Zika virus infection in humanized mice. co-infection of dengue, chikungunya and Zika in the Americas. EBioMedicine 2017; published online October 6. DOI:10.1016/j. Travel Med Infect Dis 2016; 14: 177–79. ebiom.2017.10.006. 46 Zhao H, Fernandez E, Dowd KA, et al. Structural basis of Zika 67 Tripp RA, Ross TM. Development of a Zika vaccine. virus-specific antibody protection. Cell 2016; 166: 1016–27. Expert Rev Vaccines 2016; 15: 1083–85. 47 Larocca RA, Abbink P, Peron JP, et al. Vaccine protection against 68 WHO. Current Zika product pipeline. 2016. http://www.who.int/ Zika virus from Brazil. Nature 2016; 536: 474–78. blueprint/priority-diseases/key-action/zika-rd-pipeline.pdf?ua=1 48 Rothman AL. Immunity to dengue virus: a tale of original antigenic (accessed Oct 5, 2017). sin and tropical cytokine storms. Nat Rev Immunol 2011; 11: 532–43. 69 Lazear HM, Diamond MS. Zika virus: new clinical syndromes and 49 Gulati U, Kumari K, Wu W, Keitel WA, Air GM. Amount and its emergence in the Western hemisphere. J Virol 2016; 90: 4864–75. avidity of serum antibodies against native glycoproteins and 70 Durbin AP. Vaccine development for Zika virus—timelines and denatured virus after repeated influenza whole-virus vaccination. strategies. Semin Reprod Med 2016; 34: 299–304. Vaccine 2005; 23: 1414–25. 71 Abbasi J. Zika vaccine enters clinical trials. JAMA 2016; 316: 1249. 50 Kim JH, Skountzou I, Compans R, Jacob J. Original antigenic sin 72 Pardi N, Hogan MJ, Pelc RS, et al. Zika virus protection by a single responses to influenza viruses. J Immunol 2009; 183: 3294–301. low-dose nucleoside-modified mRNA vaccination. Nature 2017; 51 Kim JH, Davis WG, Sambhara S, Jacob J. Strategies to alleviate 543: 248–51. original antigenic sin responses to influenza viruses. 73 Dyer O. Trials of Zika vaccine are set to begin in North America. Proc Natl Acad Sci USA 2012; 109: 13751–56. BMJ 2016; 353: i3588. 52 Dowd KA, DeMaso CR, Pelc RS, et al. Broadly neutralizing activity 74 National Institute of Allergy and Infectious Diseases. NIH begins of Zika virus-immune sera identifies a single viral serotype. testing investigational Zika vaccine in humans. Aug 3, 2016. Cell Rep 2016; 16: 1485–91. https://www.niaid.nih.gov/news-events/nih-begins-testing- 53 Osuna CE, Lim SY, Deleage C, et al. Zika viral dynamics and investigational-zika-vaccine-humans (accessed Oct 5, 2017). shedding in rhesus and cynomolgus macaques. Nat Med 2016; 75 Inovio. Inovio launches Zika in midst of Puerto Rico 22: 1448–55. Epidemic to explore early signals of vaccine efficacy. 2016. 54 Aliota MT, Dudley DM, Newman CM, et al. Heterologous http://ir.inovio.com/news/news-releases/news-releases- protection against Asian Zika virus challenge in rhesus macaques. details/2016/Inovio-Launches-Zika-Vaccine-Trial-in-Midst-of-Puerto- PLoS Negl Trop Dis 2016; 10: e0005168. Rico-Epidemic-to-Explore-Early-Signals-of-Vaccine-Efficacy/default. 55 Dick GW. Zika virus. II. Pathogenicity and physical properties. aspx (accessed Oct 5, 2017). Trans R Soc Trop Med Hyg 1952; 46: 521–34. 76 Kim E, Erdos G, Huang S, Kenniston T, Falo LD Jr, Gambotto A. 56 Bell TM, Field EJ, Narang HK. Zika virus infection of the central Preventative vaccines for Zika virus outbreak: preliminary nervous system of mice. Arch Gesamte Virusforsch 1971; 35: 183–93. evaluation. EBioMedicine 2016; 13: 315–20. 57 Cugola FR, Fernandes IR, Russo FB, et al. The Brazilian Zika virus 77 Omer SB, Beigi RH. Pregnancy in the time of Zika: addressing strain causes birth defects in experimental models. Nature 2016; barriers for developing vaccines and other measures for pregnant 534: 267–71. women. JAMA 2016; 315: 1227–28. 58 Wu KY, Zuo GL, Li XF, et al. Vertical transmission of Zika virus 78 Fulton TR, Narayanan D, Bonhoeffer J, Ortiz JR, Lambach P, targeting the radial glial cells affects cortex development of Omer SB. A systematic review of adverse events following offspring mice. Cell Res 2016; 26: 645–54. during pregnancy and the newborn period. Vaccine 2015; 33: 6453–65. 59 Dowall SD, Graham VA, Rayner E, et al. A susceptible mouse model for Zika virus infection. PLoS Negl Trop Dis 2016; 79 Palacios R, Poland GA, Kalil J. Another emerging arbovirus, 10: e0004658. another emerging vaccine: targeting Zika virus. Vaccine 2016; 34: 2291–93. 60 Aliota MT, Caine EA, Walker EC, Larkin KE, Camacho E, Osorio JE. Characterization of lethal Zika virus infection in AG129 mice. 80 Fernandez E, Diamond MS. Vaccination strategies against Zika PLoS Negl Trop Dis 2016; 10: e0004682. virus. Curr Opin Virol 2017; 23: 59–67. 61 Rossi SL, Tesh RB, Azar SR, et al. Characterization of a novel 81 Vannice KS, Giersing BK, Kaslow DC, et al. Meeting report: murine model to study Zika virus. Am J Trop Med Hyg 2016; WHO consultation on considerations for regulatory expectations of 94: 1362–69. Zika virus vaccines for use during an emergency. Vaccine 2016; published online Dec 1. DOI:10.1016/j.vaccine.2016.10.034. 62 Lazear HM, Govero J, Smith AM, et al. A mouse model of Zika virus pathogenesis. Cell Host Microbe 2016; 19: 720–30. 82 Plotkin SA, Mahmoud AA, Farrar J. Establishing a Global Vaccine-Development Fund. N Engl J Med 2015; 373: 297–300. 63 Dudley DM, Aliota MT, Mohr EL, et al. A rhesus macaque model of Asian-lineage Zika virus infection. Nat Commun 2016; 7: 12204.

www.thelancet.com/infection Vol 18 July 2018 e219