Rhabdoviruses as vaccine platforms for infectious disease and cancer

Abstract

The family Rhabdoviridae (RV) comprises a large, genetically diverse collection of single-stranded, negative sense RNA viruses from the order Mononegavirales. Several RV members are being developed as live-attenuated vaccine vectors for the prevention or treatment of infectious disease and cancer. These include the prototype recombinant Vesicular Stomatitis Virus (rVSV) and the more recently developed recombinant Maraba Virus, both species within the genus Vesiculoviridae. A relatively strong safety profile in humans, robust immunogenicity and genetic malleability are key features that make the RV family attractive vaccine platforms. Currently, the rVSV vector is in preclinical development for vaccination against numerous high-priority infectious diseases, with clinical evaluation underway for HIV/AIDS and Ebola virus disease. Indeed, the success of the rVSV-ZEBOV vaccine during the 2014-15 Ebola virus outbreak in West Africa highlights the therapeutic potential of rVSV as a vaccine vector for acute, life-threatening viral illnesses. The rVSV and rMaraba platforms are also being tested as ‘oncolytic’ cancer vaccines in a series of phase 1-2 clinical trials, after being proven effective at eliciting immune-mediated tumour regression in preclinical mouse models. In this review, we discuss the biological and genetic features that make RVs attractive vaccine platforms and the development and ongoing testing of rVSV and rMaraba strains as vaccine vectors for infectious disease and cancer.

Keywords:

• Rhabdovirus
• vesicular stomatitis virus (VSV)
• vaccines
• infectious disease
• cancer

Vaccines for infectious disease and cancer

Infectious disease and cancer are significant causes of morbidity and mortality amongst humans. According to the World Health Organization (WHO), infection alone was responsible for over 15 million deaths in 2010, making it the leading cause of mortality in low-income countries (Dye, 2014). The National Cancer Institute reported the burden of cancer in the United States to be 595,690 deaths in 2016, making it the second leading cause of mortality in the US (and other high-income countries), behind only heart disease (National Cancer Institute, 2017). While significant progress has been made in preventing and treating some types of infection and cancer, the long-term prognosis for others remains very poor. New treatments for these illnesses are urgently needed.

Vaccines are biological preparations designed to elicit acquired immunity against foreign entities. They have long been used to prevent and treat infectious disease. Indeed, their efficacy against foreign microbes such as polio, rabies, measles, mumps and rubella virus has been so remarkable that the vaccine is considered amongst the singular achievements of modern medicine (Watson, Hadler, Dykewicz, Reef, & Phillips, 1998). However, some microorganisms are not well controlled with our current arsenal of vaccines. These include established pathogens such as human immunodeficiency virus (HIV) as well as emerging viral threats such as Marburg, Dengue and – until recently – Ebola virus. These microbes continue to inflict significant disease burden in many countries. Further, their potential weaponization represents a global threat. Developing prophylactic and therapeutic vaccines against them is an important worldwide priority.

Vaccines have also been under development for the treatment of cancer (Mohammed, Bakshi, Chaudri, Akhter, & Akhtar, 2016). Unlike those for infectious disease, however, cancer vaccines have not yet proven effective in humans. The concept of the cancer vaccine stems from discoveries that cancer cells can be recognized as foreign by the immune system [reviewed in {Klein:1966}]. This inspired the development and evaluation of whole cell, dendritic cell, peptide and viral vaccines for treating cancer. While these vaccines have largely failed in human patients thus far, a growing appreciation of the cancer–immunity relationship has spurred the development of next-generation vectors that engage a more robust immune response towards tumours. Together with other members of a burgeoning class of cancer immunotherapies, cancer vaccines are predicted to play an important role in managing malignancy in the years ahead.

Rhabdoviruses as vaccine vectors

Replication-competent viruses are potent vaccine platforms as they generate diverse, robust and long-lived immune responses [reviewed in (Robert-Guroff, 2007)]. One virus family showing promise as a vaccine vector for both infectious disease and cancer is Rhabdoviridae (RV) [reviewed in (Rose & Clarke, 2015)]. RVs are a large family of membrane-enveloped, negative sense, non-segmented, single-stranded RNA viruses that have tremendous genetic diversity across six defined genera. While some are human pathogens (e.g. Rabies virus), many are not (e.g. Vesicular stomatitis virus, VSV; Maraba virus) – with plants, insects, horses, cows and other wildlife being their natural hosts.

The RV genome is simple and usually encodes five genes: nucleocapsid (N), phosphoprotein (P), matrix protein (M), glycoprotein (G) and large polymerase (L). These genes are expressed from a single 3’ promoter as distinct transcriptional units on a gradient from highest (N; position 1) to lowest (L; position 5) level of expression. Many RVs are pantropic (e.g. VSV, Maraba), entering cells through receptor-mediated endocytosis and replicating entirely within the cytoplasm. Most grow very fast and express their gene products to high levels.

Numerous features have provided a strong rationale for exploring RVs as vaccine vectors. First, they are highly immunogenic, eliciting strong humoral and cellular immune responses towards expressed foreign antigens while simultaneously functioning as a potent immune system adjuvant (Bridle et al., 2009; Kim et al., 2017). Equally important, they can be manipulated using reverse genetics (Lawson, Stillman, Whitt, & Rose, 1995). This enables the creation of recombinant RV strains that are either pseudotyped with pathogen glycoproteins (e.g. rVSV with Ebola GP-protein) or that deliver antigen transgenes to immune cells in secondary lymphoid organs (SLOs; e.g. rMaraba expressing the tumour-associated antigen [TAA] Mage-A3). Moreover, the structural and regulatory nature of the RV genome allows for transgene insertion upstream of the N gene, between genes, or downstream of the L gene. In theory, this allows their expression to be tuned to an optimal level. Finally, with a cloning capacity upwards of 6 kb, multiple antigens can be expressed from a single, multivalent RV vaccine.

Reverse genetics also enables the engineering of attenuated RV strains that are safer than the wildtype virus. This is especially important for neurotropic RVs such as VSV. Over the past two decades, numerous attenuation strategies have been evaluated in preclinical model systems, including non-human primates (NHPs) [reviewed in (Clarke et al., 2016)]. These include limiting viral replication to a single-cycle using pseudotyped G-deleted viruses (Publicover, Ramsburg, & Rose, 2005); decreasing the expression of viral genes (and thereby attenuating virus production) by engineering them further down the transcriptional gradient (e.g. moving the N gene from position 1 to 4 in the genome)(Ball, Pringle, Flanagan, Perepelitsa, & Wertz, 1999; Wertz, Perepelitsa, & Ball, 1998); impeding virus maturation (and thereby attenuating budding of mature virions) by truncating the C terminus of the G protein (e.g. from 29 to 1 amino acids)(Publicover, Ramsburg, & Rose, 2004); allowing for a swifter host antiviral response to attenuate virus replication, production and spread, by mutating the M protein (e.g. deleting amino acid 51)(Stojdl et al., 2003). In some cases, highly attenuated strains incorporating multiple genetic mutations have been generated, such as the rVSVN4CT1 vector that harbours both an N gene translocation and a truncated G protein (Clarke et al., 2007). Importantly, as RVs are not known to undergo homologous recombination (Pringle et al., 1981), attenuating deletions and gene rearrangements are thought to be fixed within the virus’ genome.

Several other features also make RVs attractive vaccine platforms: (i) most are insect or animal viruses, with transmission to human populations and subsequent seroconversion uncommon. As such, pre-existing neutralizing antibodies are rarely found in humans; (ii) the RV family is large and diverse. If multiple vaccine doses are required, antigenically distinct strains can be harnessed in a heterologous ‘prime-boost’ vaccine regimen; (iii) the RV life cycle is entirely cytoplasmic, which provides little opportunity for viral integration into host genomes. These features, together with a capacity to grow and purify RVs to high titers under good manufacturing practice, have led to the development and study of RVs as vaccine platforms for the past two decades.

RV vaccines for infectious disease

While numerous RV family members have been studied as vaccine vectors for infectious disease, in this review we will focus exclusively on the clinically advanced rVSV strains. rVSV has been proven an effective vector system for vaccinating against over a dozen pathogen-mediated diseases in animal model systems, including Influenza virus (Ryder et al., 2015), Zika virus (Betancourt, de Queiroz, Xia, Ahn, & Barber, 2017), Lassa virus (Safronetz et al., 2015), Andes virus (Prescott, DeBuysscher, Brown, & Feldmann, 2014), Coxsackievirus B3 (Wu, Fan, Yue, Xiong, & Dong, 2014), Respiratory Syncytial virus (Eyles et al., 2013), tuberculosis (Zhang, Dong, & Xiong, 2017), Nipah virus (Lo et al., 2014) and Chikungunya virus (Chattopadhyay, Wang, Seymour, Weaver, & Rose, 2013). However, the most extensive efforts have been directed towards developing safe and effective rVSV-vectored vaccines against HIV/AIDS and Ebola virus disease (EVD). Many preclinical studies have shown both pre- and post-exposure protection against HIV/AIDS and EVD. As such, beginning in 2011 a series of clinical trials have been conducted in which rVSV vaccines have been determined to be safe, generally well tolerated and immunogenic in humans. Indeed, during the 2014/15 Ebola virus outbreak, a ring-vaccination strategy using the rVSV-vectored Ebola vaccine (rVSV-ZEBOV) was shown to be 100% effective at preventing Ebola virus infection by 10 days post-vaccination (Henao-Restrepo et al., 2017). Buoyed by this important result, the rVSV platform is now being intensely studied for vaccination against emerging and outbreak-prone pathogens, with more than a dozen trials ongoing or recently completed for HIV and Ebola virus alone (www.clintrial.gov).

rVSV vaccines for HIV

Nearly 40 million people are infected with HIV worldwide, with ~2 million new infections occurring each year (UNAIDS. GLOBAL AIDS Update 2016, 2017). While numerous prophylactic approaches and antiretroviral drugs have proven effective in preventing and treating HIV/AIDS, a successful vaccine remains elusive (Cohen & Frahm, 2017). Indeed, since the discovery of HIV in 1983, only four vaccine strategies have even been tested for efficacy in humans. Of these, only the RV144 ‘Thai Trial’ evaluating a live Canarypox vaccine followed sequentially by a gp120 subunit boost showed even minimal protection against HIV (Rerks-Ngarm et al., 2009). Owing to its relatively low cost, ease of delivery and likelihood for high rate of adoption, an effective HIV vaccine remains the best hope for ending the HIV pandemic.

Investigations into rVSV as a vaccine vector for HIV/AIDS began two decades ago. Early studies showed that rVSV encoding the HIV env and gag genes was immunogenic and could protect mice against HIV challenge when delivered prophylactically using a homologous prime-boost vaccination regimen. The approach could be improved upon if the priming and boosting vectors were pseudotyped with glycoproteins from different VSV serotypes (so-called ‘glycoprotein exchange’ vectors), to avoid antibody-mediated vector neutralization (Ramsburg et al., 2004; Rose et al., 2001; Rose, Roberts, Buonocore, & Rose, 2000; Schell et al., 2009). Other studies found that genetically attenuated VSV vaccines such as G-truncated or -deleted mutants could also induce strong antibody and cell-mediated immunity towards HIV antigens (Publicover et al., 2004, 2005). The most effective approach, however, used heterologous prime-boost strategies to maximize T cell activation against HIV. For example, macaques primed with VSV expressing gag, pol and env and subsequently boosted with modified vaccinia virus Ankara (MVA) expressing the same antigens were completely protected from HIV challenge for > 5 years (Ramsburg et al., 2004; Schell et al., 2009). A similar approach using Simliki Forest virus as a boosting vector also generated sterilizing immunity against HIV (Schell et al., 2011).

Despite these promising results, the transition from preclinical evaluation to human trial was delayed because safety tests in NHPs showed unacceptable toxicity for early generation rVSV strains [reviewed in Clarke et al. (2016)]. Careful histopathologic examination of macaque brains after intrathalamic inoculation of rVSV strains harbouring a single attenuating mutation showed evidence of neuropathology even though those viruses did not cause overt clinical disease (Johnson et al., 2007). This prompted the development of engineered strains harbouring multiple attenuating mutations and their evaluation for vector-associated toxicity and immunogenicity (Clarke et al., 2007, 2014; Cooper et al., 2008). These efforts led to two important discoveries. The first was that attenuating mutations sometimes function synergistically to slow viral replication and promote safety. The second was that highly attenuated rVSV strains are as immunogenic as their replication-competent counterparts, at least when delivered intra-muscularly (IM). Indeed, immunogenicity after IM inoculation was found to be more tightly associated with antigen expression level than viral replication. As such, attenuated rVSV strains expressing gag antigen from the first transcription unit (to achieve maximal expression) elicited robust immune responses towards gag and were safe due to multiple attenuating mutations. When considered alongside the abundant preclinical efficacy data, these findings prompted regulatory approval for a first-in-man rVSV-based HIV vaccine trial in 2011.

This trial tested the safety and immunogenicity of dose-escalated rVSVN4CT1gag1 delivered IM into healthy volunteers (Fuchs et al., 2015). The vaccine was found to be safe, with only mild-moderate adverse events such as fever/chills and no evidence of viral shedding into blood, urine or saliva. At the highest dose, CD4⁺ T cell responses and gag antibodies were detected in 63% and one third of the participants, respectively, providing evidence of immunogenicity. A second trial in 2015 tested the same vaccine as a boosting agent, delivered IM into healthy participants six months after they were primed with an HIV multi-antigen (MAG) DNA vaccine ± IL-12 adjuvant (Li et al., 2017). This study, published recently, confirmed the safety and tolerability of the rVSVN4CT1gag1 vaccine. It also reported a significant increase in polyfunctional CD4⁺ and CD8⁺ T cell responses against gag, identifying a median of four epitopes targeted per responder. While response magnitude decreased within the year following vaccination, response rates were still maintained. Taken together, these studies demonstrated the safety of the rVSVN4CT1gag1 vaccine and provided evidence of its immunogenicity, particularly when the vaccine was used as a boosting agent. Two additional strategies are currently being assessed in clinical trials: (i) an rVSVN4CT1envC vaccine, being tested in the same prime-boost design described for rVSVN4CT1gag1 (NCT02654080); (ii) a MAG DNA prime plus rVSVN4CT1gag1 boost strategy, being evaluated in HIV-infected patients receiving combination antiretroviral drugs (NCT01859325). Both trials are scheduled to be completed mid-2019.

rVSV vaccine for Ebola virus

Discovered in Africa in 1976, Ebola virus is an extremely virulent and pathogenic virus within the family Filoviridae that causes haemorrhagic fever and can have a mortality rate as high as 90% [reviewed in (Khalafallah, Aboshady, Moawed, & Ramadan, 2017)]. Ebola virus outbreaks have occurred sporadically over the past 40 years and until recently have been controlled by isolation and containment. However, the recent outbreak in West Africa infected over 28,652 people – largely within Africa, but also outside its borders – and caused an estimated 11,325 deaths (Centers for Disease Control, 2017). The epidemic reinvigorated efforts to develop and test Ebola virus vaccines, widely considered the best potential countermeasure against its spread. Of the conventional and vector-based strategies developed, three have now been tested for safety and immunogenicity in human clinical trials, with several others under evaluation [reviewed in (Sharma, Jangid, & Anuradha, 2017)]. One of these is a live-attenuated adenovirus-vectored vaccine (chAd3-EBO-Z), which has been proven safe and immunogenic, and was recently approved for use in China (FiercePharma, 2017; Kennedy et al., 2017). Another is a live-attenuated rVSV (rVSV-ZEBOV), which was also recently approved for use, in the Democratic Republic of Congo (Agnandji et al., 2017; Henao-Restrepo et al., 2017; Huttner et al., 2015; Kennedy et al., 2017; Maxmen, 2017). This vaccine has been proven safe, immunogenic and most importantly effective at protecting against viral spread. In fact, the experimental deployment of rVSV-ZEBOV during the 2014/15 outbreak is thought to have been critical to controlling its spread.

Developed originally by the Public Health Agency of Canada, rVSV-ZEBOV comprises a wildtype rVSV backbone pseudotyped with Ebola virus GP (Garbutt et al., 2004). Preclinical studies in mice, guinea pigs and hamsters demonstrated that rVSV-ZEBOV was safe and could generate durable immune reactivity and protection against pathogenic Ebola virus inoculated after or before vaccine administration [reviewed in (Geisbert & Feldmann, 2011)]. In NHPs, vaccination 7 or 31 days prior to high-dose Ebola virus challenge completed protected animals from disease, whereas vaccination between three days pre- and one day post-exposure was partially protective (Feldmann et al., 2007; Jones et al., 2005; Marzi et al., 2015, 2016). Although the immunological correlates to protection have not been well-defined, rapid innate immunity as well as neutralizing antibody production are thought to play the major roles (Khurana et al., 2016; Marzi et al., 2013, 2016).

The preclinical success of rVSV-ZEBOV prompted its accelerated clinical evaluation during the recent outbreak. An initial placebo-controlled, double-blind phase 1/2 trial in Geneva was reported in 2015, which determined the rVSV-ZEBOV vaccine to be immunogenic; however, significant vector-associated reactogenicity was also found, from arthritis and dermatitis to vasculitis (Huttner et al., 2015). At least three other trials have since been reported, each conducted in West Africa. Two of these tracked vaccine tolerability and safety while measuring the antibody response to Ebola virus antigens (Agnandji et al., 2017; Kennedy et al., 2017). Similar to the Geneva trial, these studies found that rVSV-ZEBOV generated high antibody titres within a week post-vaccination, which were partially maintained for 6–12 months. However, in contrast to the Geneva trial, there were few serious adverse events, including no evidence of arthritis. A third open-label, cluster-randomized, ring vaccination trial also reported that rVSV-ZEBOV was safe, well tolerated and immunogenic (Henao-Restrepo et al., 2017). In addition, this trial reported for the first time on vaccine efficacy. A single IM dose of 2 × 10⁷ plaque-forming units (PFU) delivered to individuals who were in direct contact with an Ebola-infected patient (i.e. ‘contacts’) and individuals in direct contact with those contacts (i.e. ‘contacts of contacts’) was found to be 100% effective at protecting against Ebola virus infection if given as soon as possible. Nearly a dozen other clinical trials with rVSV-ZEBOV are either ongoing or have recently been completed (www.clintrials.gov).

The toxicities observed in the Geneva trial, while not reproduced in the African trials, have been ascribed to direct replication of the VSV vector in tissues such as the joints. This prompted the generation of an Ebola virus vaccine using the highly attenuated rVSVN4CT1 strain (Matassov et al., 2015; Mire et al., 2015). In contrast to the GP-pseudotyped rVSV-ZEBOV, this vaccine expresses the Ebola GP antigen from the first position of rVSVN4CT1 to achieve high transgene expression from the supra-attenuated vector. In a series of preclinical studies, the rVSVN4CT1 vaccine was reported to be safe, generate robust neutralizing antibodies towards Ebola virus antigens, and provide single-dose protection from lethal Ebola virus challenge in mice, guinea pigs and monkeys (Matassov et al., 2015; Mire et al., 2015). These encouraging results set the stage for a phase 1 dose escalation clinical trial initiated in 2015. The results from this trial – in particular how they compare to those generated with rVSV-ZEBOV – are eagerly awaited.

RV vaccines for cancer

While vaccination is a proven strategy to protect against infectious disease, efforts to develop cancer vaccines have been met with more difficulty [reviewed in (Mohammed et al., 2016)]. The reasons for this are multifaceted, but stem predominantly from differences in disease development. Infectious diseases are caused by foreign microbes, which are highly antigenic and more easily targeted by the host immune system. While 10-20% of tumours are associated with viral infections, most cancers arise from DNA mutations and/or transcriptional dysregulation. Although the genetic and epigenetic etiology of cancer generates antigens recognizable by the immune system, such as neoantigens derived from nonsynonymous DNA mutations and tumour-associated antigens (TAAs) as a product of aberrant gene expression, they are relatively weak immunogens and often exist within a highly immunosuppressed tumour microenvironment (TME) (Coulie, Van den Eynde, van der Bruggen, & Boon, 2014). As such, the quality and effectiveness of a T-cell response to antigens within tumours is generally lower than to a foreign antigen. That being said, the success of immune checkpoint inhibitors at unleashing the immune system against tumours has proven that T-cells are capable of curing human cancer if appropriately harnessed (Wilson, Evans, Fraser, & Nibbs, 2017). This breakthrough has generated significantly renewed interest in the cancer vaccine. It has also illuminated a path towards heightened vaccine efficacy via combination therapy with checkpoint inhibitors.

Therapeutic RV vaccines for cancer

Multiple clinical trials for rVSV and rMaraba vaccine vectors are currently ongoing with interim results anticipated in 2018. RV vaccines for cancer are similar to those for infectious disease in that they deliver expressed antigens to secondary lymphoid organs for T and B lymphocyte activation and function as immune system adjuvants. However, they differ because they also infect tumour cells. RV cancer vaccines have been engineered onto ‘oncolytic virus’ platforms, such as VSV^IFNβ and Maraba^MG1, which harbour attenuating transgenes and mutations that promote their infection of tumours (Brun et al., 2010; Obuchi, Fernandez, & Barber, 2003). As a cancer therapy, the oncolytic nature of these vaccines may contribute to their efficacy by killing tumour cells directly – perhaps in an immunogenic fashion – and generating inflammation to help recruit T-cells and promote T-cell function within the TME [reviewed in (Russell, Peng, & Bell, 2012)]. Indeed, early studies using strong tumour antigens have shown great promise for oncolytic RV vaccines. For example, an rVSV vaccine engineered to express human dopachrome tautomerase (hDCT) engendered T-cell mediated clearance of DCT-expressing B16 melanoma tumours in mice (Bridle et al., 2009). Similar findings have been reported for rMaraba vaccine vectors (Pol et al., 2014). These proof-of principle studies have laid the foundation for the development of RV vaccines engineered against more clinically relevant, endogenous tumour antigens.

RV vaccines targeting oncoviral antigens

Some tumours can be vaccinated against oncoviral antigens, which are encoded by oncogenic viruses that have integrated into the host genome. Being foreign antigens expressed selectively from the cancer cell, oncoviral antigens represent ideal targets for T-cell based immunotherapies. As such, rVSV and rMaraba vaccines encoding the E6 and/or E7 early proteins of human papilloma virus (HPV) have been developed to treat HPV⁺ tumours, the etiological agent of most cervical and head and neck cancers. These vaccines have been quite successful in preclinical tumour models: for instance, therapeutic vaccination of papilloma with E6-expressing rVSV caused rapid tumour clearance in rabbits (Brandsma et al., 2007). Similarly, rMaraba expressing E6/E7 antigens promoted T-cell dependent tumour regression of HPV⁺ mouse tumours when delivered as a boosting vaccine (Atherton et al., 2017). These studies demonstrate the potential of RV vaccines to treat HPV⁺ cancers and have provided justification for clinical trials testing rMaraba-E6/E7 and VSV-E6/E7 in patients with HPV⁺ solid tumours, scheduled to begin enrollment in 2018 (personal communication, John Bell). They also support the development of RV vaccines against other oncoviral antigens.

RV vaccines targeting neoantigens

Neoantigens are created when normal genes acquire nonsynonymous mutations that are immunogenic. As most tumours are of mutational origin, neoantigens represent an important reservoir of potential immunogens for cancer vaccine development. Indeed, the strong association between mutational burden and response to immune checkpoint inhibitors highlights the enormous potential of neoantigen targeting to treat human cancer (Goodman et al., 2017). Developing neoantigen-based vaccines has been made feasible by advancements in deep sequencing technology and computation algorithms that allow for the rapid identification of neoantigens within tumour genomes (Linnemann et al., 2015). In theory, this approach enables the generation of highly personalized RV vaccines targeting patient-specific mutations. While proof-of-concept RV vaccines targeting neoantigens in mouse tumours have been developed (personal communication, John Bell), their activity has not yet been reported. However, peptide-based neoantigen vaccines including personalized, multivalent vaccines have been constructed, studied and are now in clinical trial (Kreiter, Castle, Türeci, & Sahin, 2012). The genetic pliability, immunogenicity and adjuvant functioning of RV vectors provide a strong rationale for their development as neoantigen vaccine platforms.

RV vaccines targeting TAAs

TAAs are proteins whose expression is relatively restricted to malignant cells, through over- or ectopic-expression of developmental/differentiation genes. Vaccinating against TAAs is challenging, however, because TAAs are self-antigens subject to negative selection and peripheral tolerance. As a result, high-affinity TAA-specific T cells are rare and heavily regulated by tolerance mechanisms. To address these challenges, a heterologous prime-boost vaccination strategy has been developed in which the immune system is first primed with a TAA-expressing adenovirus and subsequently boosted with an RV vaccine expressing the same TAA [reviewed in Bridle, Hanson, and Lichty (2010)]. In this approach, the RV vaccine is designed to elicit a strong T-cell recall response while simultaneously infecting and killing tumour cells and conditioning the TME for enhanced T-cell function. A similar approach has been described by first priming the patient with an adoptive T cell transfer (Rommelfanger et al., 2012). Preclinical studies have shown efficacy for the approach in numerous mouse models of cancer, using either rVSV or rMaraba vectors (Bridle et al., 2009; Bridle, Stephenson, et al., 2010; Ilett et al., 2017; Pol et al., 2014). For example, an rMaraba vaccine expressing the melanoma antigen DCT generated robust anti-DCT T cell responses after adeno-DCT prime, leading to durable cancer remission in mice (Pol et al., 2014). Importantly, both the frequency and functionality of anticancer T cells is enhanced by an RV boost, including the expansion of T-cells with a central memory phenotype (Bridle et al., 2013). The success of these preclinical studies has prompted the evaluation of rMaraba boosting vaccines in human clinical trials. Indeed, an rMaraba expressing the cancer testis antigen MAGE-A3 is well into a Phase I/II clinical trial in patients with treatment-refractory, MAGE-A3 expressing solid tumours (NCT02285816; Table 1). A similar trial, also incorporating αPD-1 therapy, has been initiated in patients with lung cancer (NCT02879760).

Table 1. Rhabdovirus-vectored vaccines in clinical trial for infectious disease or cancer.

Disease	Vaccine name	Vaccine structure	Trial identifier(s)	Notes
HIV/AIDS	rVSVN4CT1gag		NCT01438606	Phase 1 trials examining safety and immunogenicity of rVSVN4CT1gag (± MAG DNA vaccine prime) in uninfected and HIVinfected individuals ± antiretroviral therapy
			NCT01578889
			NCT01859325
	rVSVN4CT1env	n/a	NCT02654080	Phase 1 trial examining safety and tolerability of rVSVN4CT1env boosting vaccine after HIV-1 nef/tat/vif, env pDNA vaccine prime
Ebola virus disease	rVSV-ZEBOV		https://clinicaltrials.gov/ct2/results?cond=ebola&term=VSV+&cntry1=&state1=&Search=Search	Phase 1–3 trials examaning safety, tolerability, immunogenicity, effectiveness, and/or durability of rVSV-ZEBOV vaccination in healthy individuals, sometimes at increased risk of Ebola virus infection
	rVSVN4CT1-EBOVGP1		NCT02718469	Phase 1 trial examining safety and immunogenicity of rVSVN4CT1-EBOVGP1 in healthy individuals
Cancer	rMVMG1mage-a3		NCT02285816	Phase 1–2 trials in solid tumour patients with MAGE-A3 expression. rMVMG1magea3 delivered IV as a boosting vector 14 and 17 days after priming with Adenovirus-MAGE-A3 vaccine (IM) ± PD1 therapy
			NCT02879760

The future of RV vaccines

After two decades of preclinical study, numerous RV vaccines are now being testing in humans for safety and efficacy against EVD, HIV/AIDs and cancer. The impressive results of rVSV-ZEBOV at preventing Ebola virus infection in West Africa assures a place for RV vectors in the vaccine armamentarium of the future. Indeed, RV vectors will likely play an important role in preventing against and treating other acute illnesses caused by emerging pathogens such as Marburg, Dengue and pathogenic RVs such as Chandipura virus. Their prophylactic or therapeutic potential against chronic illnesses such as HIV/AIDS or cancer, however, remains to be determined. The preclinical data look promising, and results from the first phase II efficacy trials should be reported within a couple of years. Going forward, it will be important to determine which RV vectors are the most immunogenic, which attenuation strategies are the safest yet sufficiently active to promote maximal vaccine activity, how durable the immune responses generated by RV vaccines are, whether unforeseen sequelae of RV-mediated vaccination emerge over time, and the precise molecular mechanisms by which RV vectors interact with the host immune system and disease. Advances in viral engineering, recombinant rescue and manufacturing will also be needed to realize the potential of recombinant RVs as platforms for highly personalized cancer vaccines.

Disclosure statement

No potential conflict of interest was reported by the authors.