Smallpox was successfully eradicated through the efforts of the World Health Organization (WHO) in 1979.^1,2 However, in the last decade there has been a renewed interest in the development of next-generation smallpox vaccines due to the threat of bioterrorism and the possible emergence of other orthopoxviruses (such as monkeypox virus) as significant human pathogens. Although the smallpox vaccines used during the smallpox eradication campaign (now called first-generation vaccines) were very efficacious, they were typically propagated in the skin of calves² under conditions that did not adhere to good manufacturing practices (GMP). VACV strains widely used during the smallpox eradication campaign included the New York City Board of Health (NYCBH), Lister, and EM-63 (Table 1). These live vaccines were commonly administered by scarification with a bifurcated needle, leading to a cutaneous reaction due to local virus replication. The resulting scar at the site of inoculation, known as the “take,” has been historically accepted as the correlate of protection for smallpox.^1-3 However, these first-generation smallpox vaccines were associated with a number of adverse reactions ranging from mild (e.g., malaise, mild rash, and fever) to severe (e.g., eczema vaccinatum, progressive vaccinia, and post-vaccinial encephalitis).^2,4 Over the years, susceptibility to more severe complications was correlated with pre-existing conditions such as atopic dermatitis, immunosupression (e.g., due to HIV/AIDS and immunosuppressive therapy), and cardiac disease. Individuals with such conditions, or with contacts that have such conditions, are currently contraindicated for smallpox vaccination.^5,6

A number of second-generation vaccines have been developed focusing on sterile cell culture techniques for vaccine propagation (Table 1). For example, the Elstree-BN vaccine developed by Bavarian Nordic was grown in chick embryo fibroblasts,⁷ and the cell-cultured smallpox vaccine (CCSV) developed by DynPort Vaccine Company was derived from the NYCBH strain grown on normal diploid MRC-5 human lung cell cultures.⁸ Similarly, Acambis (now part of Sanofi Pasteur) isolated a single clone derived from the Dryvax vaccine that was grown in Vero cells and named ACAM2000.⁹ ACAM2000 was less neurovirulent than Dryvax in mice and nonhuman primates,¹⁰ provided equivalent immunogenicity in clinical trials,¹¹ and was licensed in 2007 in the US.¹² However, the overall safety profile of ACAM2000 and other second-generation smallpox vaccines is still comparable to the first-generation vaccines such as Dryvax, with similar rates of complications.¹¹

A number of highly attenuated strains of VACV have been developed and are now being considered as safer smallpox vaccine alternatives, called third-generation smallpox vaccines (Table 1). One example is modified vaccinia Ankara (MVA), that was developed by passage of VACV strain Ankara over 570 times in chick embryo fibroblasts.¹³ Inoculation with MVA leads to abortive infections in mammals and most mammalian cells, but expression of viral genes still occurs.¹⁴ This highly attenuated VACV strain has been extensively tested in humans and has sparked considerable interest because it has been demonstrated to be extremely safe.¹⁵ However, MVA propagation is limited to chick embryo fibroblasts and a few other mammalian cell lines where yield is low,¹⁴ and the immunogenicity is not as robust compared with previous smallpox vaccine generations. For example, immune responses in phase I and II clinical trials to an MVA-based vaccine developed by Bavarian Nordic (Imvamune^®) are dose-dependent and can require two immunizations to achieve immune responses similar to first-generation vaccines.^16,17

Genetic manipulation of the VACV genome has also played a role in the development of highly attenuated VACV strains. The Copenhagen strain of VACV was considered to have higher pathogenicity,² but the deletion of 18 non-essential genes led to the highly attenuated NYVAC strain that is still immunogenic.^18,19 However, NYVAC induces lower antibody responses in humans when compared with Dryvax or Lister first-generation vaccine strains, and it does not induce anti-A27 antibodies that are seen in the immune response to first-generation vaccines and can neutralize intracellular mature virus (IMV).²⁰ Two major disadvantages of highly attenuated VACV strains such as MVA and NYVAC is that they do not produce a “take” in vaccinees and their efficacy against smallpox was never determined.

An additional third-generation vaccine is LC16m8, an attenuated VACV derived from the Lister strain that has an excellent safety profile.^21-23 LC16m8 contains a mutation in the B5R gene, which causes the virus to produce smaller plaques and replicate less efficiently in Vero cells,^24,25 but unlike MVA, LC16m8 produces a “take” in vaccinees.^22,23 This vaccine was licensed in Japan and used in the 1970s eradication campaign, but since smallpox was no longer endemic at that time, its efficacy against smallpox is currently unknown. A disadvantage that must be considered is that LC16m8 does not induce neutralizing antibodies against the B5 protein, the main target for extracellular enveloped virus (EEV) neutralizing antibodies.²⁶

A number of different approaches are being used for the development of the so-called fourth-generation smallpox vaccines that eliminate the possibility of the VACV vector to cause adverse events or revert to a more pathogenic phenotype (Table 1). These include the development of subunit and DNA vaccines typically composed of VACV (or variola virus counterpart) membrane proteins that elicit neutralizing antibodies against the IMV and EEV forms of the virus.^27-33 A particularly new approach is the use of conserved and immunogenic multi-T-cell epitopes that are used in a DNA-prime, peptide boost vaccine regimen.³⁴ These fourth-generation vaccines have shown to be protective using animal models, but none are currently being tested in clinical trials, and to be efficacious against smallpox they will likely require booster immunizations.

The successful use and extensive characterization of VACV during the smallpox eradication campaign, along with the ability to genetically manipulate its large dsDNA genome while retaining infectivity, its heat stability and low cost of production, makes VACV particularly attractive for the development of animal vaccines.^35,36 A large number of antigens from animal pathogens have been expressed in VACV, and the majority elicit protective immune responses (examples shown in Table 2). The most successful recombinant VACV vaccine has been the oral vaccine for sylvatic rabies (Raboral V-RG^®) that expresses the rabies glycoprotein (G).^37,38 This recombinant virus is packaged into vaccine baits that are distributed in areas affected by rabies. The heat stability of the V-RG vaccine proved to be extremely helpful in sylvatic rabies vaccination programs, which have led to better control of rabies in Europe and North America.³⁹ Several recombinant VACV vaccines engineered to express the fusion and hemagglutinin glycoproteins of rinderpest virus have also been developed,^40,41 providing sterilizing immunity to cattle challenged with virulent rinderpest virus over one year after vaccination.⁴¹ Recombinant VACVs expressing heterologous viral antigens have also been protective against peste-des-petits-ruminants virus, vesicular stomatitis virus, Newcastle disease virus, canine distemper virus, and Rift Valley fever virus in a variety of animal species (reviewed in Table 2).

The use of VACV as a vector for human vaccines against infections agents has also been extensively investigated (Table 3), with most of the effort focused on the development of vaccines against HIV. Initially, replication-competent VACVs were typically employed, but more recent efforts center on the use of replication-defective poxviruses due to safety concerns. In addition, replication-defective poxvirus vectors, including MVA, NYVAC, and avipoxviruses (such as canarypox virus and fowlpox virus), offer the advantage of allowing multiple booster immunizations even in subjects with pre-existing immunity.⁴² The only Phase III trial to show any evidence of protection against HIV has been the RV144 trial in Thailand.⁴³ The RV144 vaccine regimen consisted of four priming injections of a recombinant canarypox expressing HIV-1 Gag, protease, and Env, followed by two booster immunizations with a recombinant Env subunit vaccine. The trial involved 16,402 subjects and the estimated vaccine efficacy (prevention of HIV-1 infection) was 31.2%.⁴³ The results from the RV144 Thai trial, albeit modest, reinvigorated the HIV vaccine community and their interest in poxvirus-vectored HIV vaccines. A number of Phase I and II HIV clinical trials, usually in the form of a DNA prime and recombinant MVA or NYVAC boost, have shown that these vaccine regimens are safe, immunogenic, and have the potential to improve the efficacy obtained with the RV144 Thai trial (Table 3).

Recombinant MVA has also been extensively used alone or in prime-boost strategies in vaccine clinical trials for other viral diseases such as influenza and hepatitis B, as well as bacterial and parasitic diseases such as tuberculosis and malaria (Table 3). Three malaria Phase I trials with recombinant chimpanzee adenovirus (ChAd63) and MVA expressing different Plasmodium falciparum antigens have shown that this prime-boost strategy is safe and immunogenic.^44-46 Likewise, Phase I trials with an MVA vector expressing the Mycobacterium tuberculosis 85A antigen, aimed to serve as a booster immunization after bacille Calmette-Guerin (BCG) vaccination, showed that the vaccine can induce potent Th1 responses.⁴⁷ Lastly, a number of other constructs are being developed and investigated in pre-clinical trials, including vaccines against hepatitis C virus,⁴⁸ respiratory syncytial virus,⁴⁹ anthrax (with the advantage of being a dual vaccine against anthrax and smallpox),⁵⁰ and Nipah virus.⁵¹

The ability of VACV to induce potent immune responses to tumor-associated antigens (TAAs) expressed in its genome has been employed for the development of immunotherapies for cancer (Table 4). One example is PROSTVAC^® (Bavarian Nordic), a therapeutic cancer vaccine for prostate cancer that consists of a replication-competent VACV prime followed by multiple replication-defective avian poxvirus (fowlpox) boosts.^52,53 Both poxviruses express a modified prostate-specific antigen (PSA), along with three T-cell costimulatory molecules termed TRICOM (B7.1, ICAM-1, and LFA-3). The immune response to the altered PSA (with a single amino acid change in an HLA-A2 epitope) is aimed to target cancerous prostate cells, while the costimulatory molecules increase the immunogenicity of the constructs. This treatment has been studied in Phase II clinical trials in patients with metastatic castration-resistant prostate cancer, where treatment was well tolerated, death rate was reduced by 44%, and the median overall survival was 8.5 mo longer than in patients receiving control vectors.^52,53 This increase in overall survival was noticeably superior to that afforded by the currently approved treatment (docetaxel, 2–3 mo increase in overall survival), and Phase III clinical trials are ongoing.^52,54 Bavarian Nordic is also performing clinical trials with an MVA vector (MVA-BN^® PRO) expressing two prostate TAAs, PSA along with prostatic acid phosphatase (PAP),^54,55 and with MVA-BN^® HER2, an MVA vector expressing the TAA human epidermal growth factor receptor 2 (HER2) to treat breast cancer.^54,56

TroVax^® (Oxford BioMedica) is an MVA vector that expresses the human oncofetal antigen 5T4, a placental glycoprotein overexpressed in a number of different cancers (Table 4).⁵⁷ When administered alone or in conjuction with other treatments (e.g., chemotherapy and cytokines), Trovax^® mounted high anti-5T4 immune responses that were associated with increased overall survival.^54,57-61 Poxvirus vectors expressing another TAA, the cancer-testis NY-ESO-1 antigen, were tested in Phase II clinical trials in a VACV prime followed by fowlpox boost regimen with melanoma and ovarian cancer patients that were at risk for recurrence or progression after completion of their primary therapy.^54,62 The trials provided evidence for an extended overall survival and progression-free survival in patients that developed detectable anti-NY-ESO-1 immune responses.⁶²

TG4010 (Transgene SA) is an MVA vector that expresses mucin-1 (MUC-1), a TAA expressed by glandular epithelia and by some cancers, as well as the immostimulatory cytokine IL-2 (Table 4).⁶³ Results from a Phase IIb study with non-small-cell lung cancer patients in combination with chemotherapy suggests that TG4010 has beneficial effects and additional trials are ongoing.^54,64 Finally, CV-301 (Bavarian Nordic), formerly known as PANVAC, is an MVA vector that in addition to MUC-1, also expresses carcinoembryonic antigen (CEA) and the TRICOM set of T-cell costimulatory molecules, having a potential effect on multiple cancers.⁶⁵ A recent pilot study in patients with metastatic breast and ovarian cancer suggests that CV-301 may be beneficial to patients and a Phase II study is underway.^54,65

Oncolytic viruses are promising new therapies for the clinical treatment of cancers. As a replication-competent virus, VACV displays a natural tumor tropism and kills cancer cells through apoptosis and other mechanisms.⁶⁶ Moreover, recombinant VACVs in which the vaccinia growth factor (VGF) and thymidine kinase (TK) genes have been deleted acquire enhanced tumor selectivity (tropism), likely because cancer cells overexpress epidermal growth factor receptor (EGFR) and have high metabolic rates that make TK expression by VACV dispensable.^67,68 Systemic administration of these oncolytic VACV vectors (e.g., intravenously) targets both tumors and metastases, and current clinical trials show that VACV vectors can be an effective oncolytic cancer therapy (Table 5). In addition, recombinant VACVs expressing host immunomodulating genes such as human GM-CSF and other cytokines, anti-angiogenic agents, and extracellular matrix genes also show enhanced tumor selectivity and oncolytic effects.⁶⁶ For example, JX-594 (Jennerex Biotherapeutics) is a TK negative VACV expressing GM-CSF. In patients with hepatocellular carcinoma, treatment with JX-594 resulted in antitumor and antivascular activities,⁶⁹ and in clinical trials patients with metastatic liver cancer and metastatic melanoma tolerated treatment and anti-tumor effects were observed.^70,71 Another example is vvDD-CDSR (Jennerex Biotherapeutics), a TK and VGF negative vector engineered to express a reporting gene (somatostatin receptor) for molecular in vivo imaging after systemic delivery, as well as the prodrug-activating enzyme cytosine deaminase for therapy with 5-fluorocytosine.^72,73 These modifications allow vvDD-CDSR to be highly tumor selective and oncolytic. Finally, GL-ONC1 (Genelux), also known as GLV-1h68, is a light-emitting oncolytic VACV expressing a luciferase-green fluorescent protein (GFP) fusion gene and containing deletions in the F14.5L, TK, and A56R genes.⁷⁴ In pre-clinical trials it showed tumor-specific replication and solid tumor size reduction.^74-76 In addition, VACVs with tumor selectivity and light-emitting or imaging features (such as expression of somatostatin receptor and ¹¹¹In-pentreteodide treatment) can be innovative and helpful tools to non-invasively detect and locate primary and metastatic tumors, as well as to track the therapeutic vector during treatment.^72,77

Since the eradication of smallpox more than 30 y ago, VACV has experienced a renaissance of interest as a viral vector for the development of recombinant vaccines, immunotherapies, and oncolytic therapies, as well as the development of next-generation smallpox vaccines. This renewed interest in both replication-competent and replication-defective VACVs has driven a number of vaccine and therapeutic candidates to clinical trials. Replication-competent VACVs are used for oncolytic therapy, and as live vaccines they elicit superior immune responses, but safety is a concern due to potential adverse events. Conversely, replication-defective VACVs are not as immunogenic as their replication-competent counterparts, but are extremely safe and offer the capacity to be administered multiple times with minimal interference from pre-existing immunity. A number of approaches have been used to enhance the safety of replication-competent VACV vectors while maintaining their immunogenicity, such as the deletion of viral genes and expression of cytokines. A new approach under development to generate next-generation smallpox vaccines that are efficacious and safer is the addition of safety mechanisms to second-generation VACV vectors that should allow the conditional replication of the virus and the production of the “take,” the correlate of protection against smallpox (Hagen, Titong, and Verardi, unpublished data). With these built-in safety systems, VACV replication can be controlled so that adverse events in vaccinees and contacts can be treated with antibiotics. These new technologies can also be applied for the development of VACV vectors for human vaccines and therapies with enhanced safety features.

A range of potential new uses is under development in areas such as tumor imaging and enhanced oncolytic and tumor selectivity. The fact that VACV has the capacity to hold at least 25 kb of heterologous DNA while retaining infectivity also enables the development of multi-pathogen, multi-epitope polyvalent vaccines. Hence, it seems inevitable that this “old” vaccine will lead to “new” uses in the near and distant future.