Modulation of immune responses using adjuvants to facilitate therapeutic vaccination

Abstract Therapeutic vaccination offers great promise as an intervention for a diversity of infectious and non‐infectious conditions. Given that most chronic health conditions are thought to have an immune component, vaccination can at least in principle be proposed as a therapeutic strategy. Understanding the nature of protective immunity is of vital importance, and the progress made in recent years in defining the nature of pathological and protective immunity for a range of diseases has provided an impetus to devise strategies to promote such responses in a targeted manner. However, in many cases, limited progress has been made in clinical adoption of such approaches. This in part results from a lack of safe and effective vaccine adjuvants that can be used to promote protective immunity and/or reduce deleterious immune responses. Although somewhat simplistic, it is possible to divide therapeutic vaccine approaches into those targeting conditions where antibody responses can mediate protection and those where the principal focus is the promotion of effector and memory cellular immunity or the reduction of damaging cellular immune responses as in the case of autoimmune diseases. Clearly, in all cases of antigen‐specific immunotherapy, the identification of protective antigens is a vital first step. There are many challenges to developing therapeutic vaccines beyond those associated with prophylactic diseases including the ongoing immune responses in patients, patient heterogeneity, and diversity in the type and stage of disease. If reproducible biomarkers can be defined, these could allow earlier diagnosis and intervention and likely increase therapeutic vaccine efficacy. Current immunomodulatory approaches related to adoptive cell transfers or passive antibody therapy are showing great promise, but these are outside the scope of this review which will focus on the potential for adjuvanted therapeutic active vaccination strategies.


| INTRODUC TI ON
Vaccines have made an enormous contribution to the reduction of morbidity and mortality across the globe. The expression "therapeutic vaccine" may sound counter intuitive since "vaccines" are traditionally used as prophylactic medicines with the aim to prevent, rather than to treat, diseases, mostly viral, or bacterial infections.
The terms "therapeutic immunization" or "therapeutic vaccines" used in this paper define vaccines, which are used therapeutically to treat a medical condition, such as an ongoing, chronic, often debilitating health problem, or an unwanted biological response.
Therapeutic vaccines aim to reprogram the immune system of the patient in order to better recognize and neutralize specific deleterious molecular targets or immune cells. 1 Hence, the vaccine could target antigens associated with an infectious (chronic) disease or non-infectious diseases such as cancer, allergy, drug (eg, nicotine) addiction, or self-molecules associated with autoimmune/autoinflammatory situations (eg, hypertension, neurological disorders, atherosclerosis, diabetes). 2  The next challenge is how to generate the desired immune response against the chosen antigen, ideally with long-term immunological memory. This requires the choice of a suitable immunostimulator that is able to induce, prolong, magnify, and steer the required immune response. This is the role for vaccine adjuvants, which with the increasing knowledge in innate immune mechanisms has moved from being "immunologists dirty little secret" to vaccinologists requisite for success. The two main objectives for the vaccine This in part results from a lack of safe and effective vaccine adjuvants that can be used to promote protective immunity and/or reduce deleterious immune responses.
Although somewhat simplistic, it is possible to divide therapeutic vaccine approaches into those targeting conditions where antibody responses can mediate protection and those where the principal focus is the promotion of effector and memory cellular immunity or the reduction of damaging cellular immune responses as in the case of autoimmune diseases. Clearly, in all cases of antigen-specific immunotherapy, the identification of protective antigens is a vital first step. There are many challenges to developing therapeutic vaccines beyond those associated with prophylactic diseases including the ongoing immune responses in patients, patient heterogeneity, and diversity in the type and stage of disease. If reproducible biomarkers can be defined, these could allow earlier diagnosis and intervention and likely increase therapeutic vaccine efficacy. Current immunomodulatory approaches related to adoptive cell transfers or passive antibody therapy are showing great promise, but these are outside the scope of this review which will focus on the potential for adjuvanted therapeutic active vaccination strategies.

K E Y W O R D S
adjuvant, autoimmunity, cancer, cellular immunity, therapeutic, vaccine adjuvants are to (a) ensure that the antigens are delivered to those cells of the innate immune system specialized in inducing the desired immune response and (b) activate the innate immune system to direct the inducible adaptive immunity in the required direction.
Modern adjuvants thus typically consist of both a delivery system, that can be composed of emulsions, liposomes, polymeric nanoparticles etc and immunostimulators that are typically ligands for socalled pattern recognition receptors (PRRs) on the innate immune cells. These ligands are typically mimicking pathogen or damage-associated molecular patterns (PAMPs/DAMPs), including Toll-like receptor (TLR) agonists, C-type lectin receptor (CLR) agonists, Nod-like receptor (NLR) agonists, and RIG-like receptor (RLR) agonists. Only a few adjuvants have been approved in vaccines for human use, including aluminum salts, Virosomes, MF59™, AS01™, AS03™, AS04™, and CpG, but many more are in clinical development and the repertoire will expand during the next decade. This subject has recently been reviewed in details in the book "Immunopotentiators in Modern Vaccines" which provides in-depth insights and overviews of the most successful adjuvants, including those that have been included in licensed products and the most promising emerging technologies. 3 Other key issues at this stage are formulation safety, stability, capacity for sterile filtration, consistency of component availability, and cost.

| Antibody-based immune responses against extracellular targets
Many molecules that are targets of monoclonal antibody therapies may also be candidates for actively elicited antibodies, triggered by therapeutic immunization. Hence, neutralizing antibodies induced by vaccination may be alternatives to, often very expensive, passively administered monoclonal antibodies. However, the duration of such active circulating antibody responses vs declining concentrations of passively administered monoclonals needs to be considered for safety reasons. In case of drug addictions, neutralizing antibodies induced by vaccines against cocaine or nicotine may be ideal; however, when acute reversibility or temporary inhibition of responses is the aim, passive delivery of monoclonal antibodies may be preferred over actively generated antibodies.
One of the main effector processes triggered by a vaccine is the production of antibodies by antigen-specific B lymphocytes. Soluble antibody molecules can specifically detect and bind their antigen target, promoting its neutralization and opsonization by phagocytes (macrophages and neutrophils). Antibodies are also able to (a) block the activity of certain toxins, (b) prevent the spread of harmful infectious agents, (c) keep viruses from entering healthy cells, and (d) activate the complement cascade, which helps in microbe clearance. B cells can directly recognize antigenic moieties of extracellular pathogens through the immunoglobulins present in their outer membrane, the so-called B cell receptors (BCR). Upon encountering the antigen, B cells become activated and differentiate into plasma cells, which produce and secrete specific IgM antibodies (without class switching) in a process known as T cell-independent activation. 4 Activation of CD4 + T lymphocytes is essential for optimal B cell priming by enhancing antigen presentation. 5 T-helper   lymphocytes were initially subdivided into two main groups with   counter-regulatory functions depending on the cytokine pattern: a Th1 subset, which participates in cell-mediated immunity and is associated with IFN-γ secretion; and a Th2 subset, which was associated with enhanced proliferation of activated B cells, promoting differentiation toward plasma cells, enhanced expression of MHC class II molecules, and isotype class switching by secreting several cytokines such as IL-4 and IL-13. 10 This dichotomy became outdated following the description of other CD4 + T-helper populations including Th9, Th17, and Th22 cells, regulatory T cells (T REG ), and T follicular helper (T FH ) cells. 11 In terms of therapeutic vaccines for treating autoimmune and inflammatory conditions and maintaining homeostasis and self-tolerance, T REG play a central role. [12][13][14] T REG constitute 5% of circulating T cells and are characterized by expression of the transcription factor Foxp3. 15 In addition to these "natural" T REG , a population of inducible type 1 T REG cells (Tr1 cells) is characterized by the secretion of IL-10 and expression of CD49b and lymphocyte activation gene 3 protein (LAG3). 16 Of particular relevance to vaccine-induced antibody responses, are the follicular T FH cells, which are specialized in providing T cell help to B cells. 17,18 Their specific location in the lymph node B cell follicle and cellular interactions allows them to play a key role in the induction and regulation of antibody production 19 and B cell memory responses. This T cell-dependent B cell activation enables antibody class switching and elicits more robust responses and higher-affinity antibodies than the T-independent activation, highlighting the importance of optimal targeting of DCs and the subsequent instruction of CD4 + T cells in the adaptive response to a vaccine. These interactions and potential adjuvant interventions are highlighted in Figure 2. Differentiation of T FH cells is mediated by STAT-3 (signal transducer and activator of transcription   3), activating cytokines, secreted from DCs (IL-6, IL-12 and IL-27),   B cells (IL-6 and probably IL-27), and CD4 + T cells (IL-21) although there are significant differences in cytokine requirements between mice and humans. 17  and medication scores. 21 The principle behind AIT is to expose the allergic individual to the specific allergen(s), leading to immunological tolerance or altered antibody responses with a lasting effect, even after termination of the treatment. 22,23 Adjuvants and adjunct therapies have, in the context of AIT, been investigated since the 1990s, with the aim to either increase the efficacy, lower the adverse events, or shorten the treatment. So far, we have not seen a major breakthrough, but several clinical trials are still aiming to change this picture ( Table 1). As previously mentioned, aluminum hydroxide (alum) has been the adjuvant of choice in SCIT.
The effect of alum can partly be attributed to its adsorptive properties, which leads to increased immunogenicity due to the slow and sustained allergen release from the alum particles 24 (depot effect).
Importantly, allergens strongly adsorbed to alum particles are less accessible to the immune system, which has been shown to decrease local side effects. 25 A number of pathways have been implicated in alum adjuvanticity, including DNA release, 26 prostaglandin E2, and cholesterol interaction. 27 The mineral salt, calcium phosphate, and the natural amino acid L-tyrosine are also used as depot adjuvants in SCIT. Both of these adjuvants enhance IgG production with a limited increase of IgE. 28 More recently, Toll-like receptor (TLR) ligands have been applied as adjuvants in SCIT. One example is monophosphoryl lipid A (MPL), which is a non-toxic derivative of lipopolysaccharide (LPS) which interacts with TLR4 on innate immune cells and can promote Th1 responses. 29 In combination with L-tyrosine, MPL is being investigated in phase III clinical studies as an adjuvant in short-course SCIT treatment of pollen-induced allergies. However, the most recent of these studies, treating birch pollen allergics, failed to meet the primary endpoint of reducing the combined symptom and medication score (Study B301). Glucopyranosyl lipid A (GLA) is similarly a TLR4 ligand, inducing a Th1 response by activating dendritic cells. 30 This adjuvant is currently being evaluated in a clinical study with SLIT treatment of peanut allergic patients (NCT03463135). Attempts have been made to improve AIT using adjuvants targeting TLRs other than TLR4. CpG oligodeoxynucleotides (CpG-ODN) resemble bacterial DNA and bind to the endosomal receptor TLR9. In animal models of asthma, it has been shown that CpG-ODN downregulate the allergic Th2 response and induce a Th1 and T-regulatory response. 31 This concept was tested by SCIT treatment of ragweed allergic patients where the antigen Amb a 1 was conjugated to a CpG-ODN. Initial results from a phase II study showed promising results, but in a later phase IIb study, the effect was not different from placebo. 28 Nanoscale adjuvants, including virus-like particles (VLP), have also been tested for efficacy in SCIT. VLPs are virus-shaped particles made of coat proteins or capsids from viruses or bacteriophages.
These particles are highly immunogenic and linking allergens to their surface enhances their immunogenicity. However, it has been demonstrated that VLP bound allergen has a strongly impaired ability to bind to surface bound IgE and to induce mast cell degranulation. 32 The enhanced immunogenicity was demonstrated in a small clinical trial, where high titers of specific antibodies were observed after immunization with the antigen Der p 1 conjugated to VLP. 33 Furthermore, VLP combined with CpG-ODN has been tested as a SCIT adjuvant in clinical trials, where it showed reduction of clinical symptoms in house-dust mite allergic patients. 34 Interestingly, this reduction could be seen both with and without coupling of allergen to the VLPs. 35 An alternative to adjuvants is co-administration of adjunct biopharmaceuticals. In several studies, AIT has been combined with omalizumab (anti-IgE antibody) for the treatment of allergic rhinitis or asthma, resulting in fewer side effects compared to AIT alone. 36 In food allergy, it has been shown that addition of omalizumab to oral immunotherapy was beneficial effect in terms of adverse events, but had no effect on tolerance induction or sustained responses. 36 Dupilumab, an anti-IL-4 receptor alpha-specific antibody, is currently being investigated as adjunct therapy to peanut AIT. Mouse studies have indicated that peanut AIT and anti-IL-4R antibody treatment show a synergistic treatment effect.

| Vaccines for substance use disorders
Addictive drugs are a chemically heterogeneous group of small molecules that are able to pass the blood-brain barrier (BBB) and target the distinct neurotransmitter systems in the brain. Vaccines have the potential to induce anti-drug antibodies that cannot cross the BBB but can potentially bind to the drug and inhibit its transport to the brain, without altering brain function. Because the addictive drugs are small non-immunogenic molecules, a hapten-carrier approach is used which requires conjugation to a carrier protein to enhance the induction of drug-specific antibodies. For example, the key components of a conjugate nicotine vaccine are: a B cell epitope, in this case nicotine; a T cell epitope, provided by the carrier protein as used in polysaccharide conjugate vaccines for capsulated bacterial pathogens and an adjuvant that enhances vaccine immunogenicity. 37 Although the vaccine principle is the same for all drugs, when designing the hapten vaccine, it is important to consider drug properties such as size, chemical structure, metabolism, and biodistribution. 38 The efficacy of induced antibodies will depend on a number of variables including the carrier, hapten density, aggregates and adducts, and a choice of adjuvant. [38][39][40][41][42][43] Nicotine, a psychostimulant, is the main addictive component in second-generation vaccine that has shown promise for nicotine vaccination. 52 The N4N hapten is a covalent modification of pyridine and has much higher nicotine affinity than 3'aminomethylnicotine from the NicVax vaccine. The N4N hapten is conjugated to flagellin but has not yet been tested clinically.
A different vaccine approach for inducing drug-specific antibody responses involves particle-based vaccines, which are built from either polymers, liposomes, peptides, virus-like particles, or other combinations. 53-55 These self-assembling particle vaccines are anticipated to enhance the activation of antigen-presenting cells (APC), to promote stronger T-helper cell responses, and to stimulate the differentiation of memory B cells. 56,57 Additionally, the hapten load can be controlled and the delivery of adjuvants and other immunomodulators to APCs made more efficient. 42 The nanoparticle-based vaccine SEL-068 from Selecta Bioscience consists of nicotine bound to the surface of polymers, a synthetic TLR ligand, and a T-cell helper peptide. In preclinical studies in non-human primates, the vaccine blocked the development of nicotine discrimination, a behavioral experimental procedure to test the effect of nicotine. 58 The Selecta group showed that codelivery of an antigen with a TLR7/8 or TLR9 agonist in synthetic polymer nanoparticles increased drug immunogenicity with minimal systemic production of inflammatory cytokines. 59 SEL-068 is currently being evaluated in phase 1 clinical trials.
Another particle-like vaccine in preclinical studies incorporates a synthesized short trimeric coiled-coil peptide (TCC) that creates a series of B and T cell epitopes with uniform stoichiometry and high density. 60 Vaccination with this antigen and alum and a TLR4 agonist

| Vaccines against chemical hazards
DDT (1,1,1-trichloro-2,2-bis (p-chlorophenyl) ethane) has been used widely as a pest control agent but was banned in 1970 because of its harmful effects on wildlife and human health. 67 This compound is therefore a chemical hazard for both human health and the environment. The biodegradation of DDT is very slow, 68 and in animals and humans, the DDT accumulates and is stored in adipose tissue. 69 When an organism is a part of the food chain, a biomagnification occurs, where the hazard is accumulated. 70 Although the liver is able to transform some of the DDT to DDE or DDD, 71 these degradation compounds are not eliminated, but are stored even more avidly. 72 Research has been conducted into the potential of the immune system to eliminate DDT following vaccination. The toxin DDT was made immunogenic by conjugating it with keyhole limpet hemocyanine (DDT-KLH). Mice were immunized subcutaneously using aluminum hydroxide adsorbed DDT-KLH conjugate, where the second group only received KLH adsorbed to aluminum hydroxide, and the third group was used as a control to provide information on DDT levels in serum in untreated animals. The mice were then fed with chow containing 40 mg/kg of DDT for 45 days. 73 The concentration of DDT and its metabolites (DDE and DDD) was analyzed in various tissues, and DDT-specific antibody titers were determined. Higher antibody responses were detected in mice vaccinated with the DDT-KLH conjugate, and DDT, DDE, and DDD levels in adipose tissue, blood, brain, and spleen were significantly reduced, compared to the group that received native unconjugated DDT. 73 This demonstrates that immunization against a chemical hazard may be used to treat animals or humans exposed to high levels of such toxic compounds, as well as prevent them from having these compounds accumulating in their body.

| Alzheimer's disease vaccines
Alzheimer's disease (AD) is a progressive neurodegenerative disorder, characterized by plaques of misfolded amyloid β (Aβ) and tau protein aggregates in neural tissue, and cerebrovascular dysfunction resulting from damaged small blood vessels in the brain, 74 which eventually leads to dementia in the elderly population. Currently, it is considered an incurable disorder with limited treatment options. The mechanism(s) underlying the cognitive decline in Alzheimer's disease still has not been clearly unraveled, which makes it difficult to judge whether current therapies target the symptoms, rather than the key molecule(s) causing the pathology. In AD transgenic animal models, immunotherapeutic targeting of B cell epitopes of (misfolded) Aβ and/or tau seemed a most logical strategy to target neural plaques or vascular aggregates. Indeed, both vaccines, and passive antibodybased interventions, resulted in a clear reduction of Aβ pathology and cognitive benefits, with limited local inflammatory adverse effects. However, despite these promising preclinical results, such approaches in human studies showed limited evidence of significant clinical benefits, even despite the postmortem observed clearance of amyloid pathology, and acceptable tolerability. [75][76][77] Several redesigned vaccines against the (modified) variants of soluble or aggregated Aβ and tau protein (with or without protein carrier Qβ or KLH, lacking common T cell epitopes) formulated with saponin QS-21, liposomes plus TLR4 agonist, or KLH-alum, as immune modulating adjuvants, are being evaluated in early-stage trials.  (Table 3). 93

BOX 1 Cancer vaccine prototypes based on TACAs targeted by antibody induction Le y AND SLe a -BASED VACCINES
A Le y -based vaccine, effective in preclinical trails, was generated by conjugating the synthetic pentasaccharide Le y to KLH and injected with QS-21 as an adjuvant. Vaccination triggered IgM and, to a lesser extent, IgG antibodies and proved to be selectively toxic to Le y -positive cells. 97 This potential vaccine entered a phase I clinical trial against ovarian cancer, but was discontinued because the antibodies induced were low affinity IgM molecules. The hexasaccharide SLe a (CA19-9) was also used as a TACA linked to KLH and co-administered with QS-21 or with the semisynthetic saponin adjuvant mixture GPI-0100. In animal tests, vaccination with SLe a -KLH alone induced moderate antibody titers, which were increased by using GPI-0100 as an adjuvant. The induced IgM and IgG antibodies mediated cytotoxicity against the SLe a -positive human adenocarcinoma cell line SW626. No cross-reactivity to other or similar carbohydrate antigens (SLe x , Le a or Le y ) was detected. 98

GLOBO H-BASED VACCINES
The first total synthesis of the Globo H hexasaccharide was reported in 1995. 99   The fucosyl-GM 1 hexasaccharide is expressed on many small-cell lung cancers but not on normal cells. The synthetic fucosyl-GM 1 linked to KLH was employed in combination with QS-21 as an adjuvant. 109 Upon vaccination, an IgM antibody response against fucosyl-GM 1 and against tumor cells expressing fucosyl-GM 1 was elicited.

Tn-, STn-, TF-, AND STF-BASED VACCINES
The MUC1-related Tn, TF, STn, and STF TACAs, expressed in more than 90% of primary adenocarcinomas, are considered pancarcinoma antigens and have been widely used to design cancer vaccines. They can be found attached through O-glycosidic linkages to serine or threonine residues of tandem repeat peptide sequences. A synthetic, Tn/TF-containing MUC1 peptide sequence conjugated to bovine serum albumin (BSA) has been developed as a vaccine candidate. A higher antibody response was obtained when a Tn or a TF residue was linked to a threonine residue (19). The same authors also synthesized the MUC1 tandem peptide glycosylated with STn and 2,6-STF antigens at a serine residue (S15), maintaining Tn or TF antigens linked to threonine (T9). After conjugation to BSA, the vaccine candidates were co-administered with complete Freund's adjuvant for the first inoculation and with incomplete

| Therapeutic cancer vaccines
Anti-tumor therapeutic vaccination should generate not only tumorspecific CD8 + T cells but also tumor-specific tissue-resident memory This vaccine was able to generate a therapeutic response due to specific immunity against MUC1 and to a non-specific anti-tumor response elicited by the adjuvant. 95 Another example of self-adjuvating potential cancer vaccine was developed by covalent conjugation of α-galactosylceramide (αGalCer), a CD1d and invariant natural killer T (iNKT) cell ligand, with sialyl Tn (STn) TACA. This vaccine showed remarkable efficacy in inducing a strong STn-specific IgG response. 111 In an effort to improve the multivalent presentation of TACAs as well as to improve anti-glycan immunity, the formulation of Tn antigen conjugated to αGalCer into liposomes was recently published. 112 Liposomes containing 1,2-diasteraroyl-sn-glycero-3-phoshpocholine (DSPC) and cholesterol were prepared by lipid extrusion and co-formulated with Tn-αGalCer glycoconjugate and αGalCer.
The combination of DSPC and cholesterol for liposomal vaccines was known to induce a strong antibody (particularly IgG1) response in mice. 113 In another vaccine design, the Tn antigen was linked to virus-like particles (VLP) of the bacteriophage Qβ, eliciting higher and more diverse antibody responses. 114 A four-component self-adjuvating multivalent cancer vaccine prototype was also proposed, relying on a cyclic peptide scaffold presenting four residues of Tn antigen and decorated with a B-cell epitope. The multivalent scaffold also

BOX 1 (Continued)
why their tumor infiltration is correlated with good clinical outcomes in many cancers. Accordingly, it appears that eliciting Trm represents a key target for the success of cancer vaccines.
The phenomenon of antigen spreading, which corresponds to the secondary CD8 + T cell response against antigen epitopes that are not present in the vaccine but likely result from local release after the first wave of tumor-specific CD8 + T cells, provides additional evidence for the important role of these effector cells. 120  shown to elicit CD8 + T cells in mice and in humans. 122 However, some discrepancies between results obtained in mice and humans, especially for DNA vaccines, have to be addressed. 123 The plasmid-based DNA vaccines were thus very efficient in many species including mice and non-human primates, but early human studies failed due to low efficacy. The plasmid technology has greatly improved over the years, both when it comes to codon optimization increasing expression of the antigens, and plasmid delivery within the body. So, it will be interesting to follow these techniques in the future.
Immunostimulatory adjuvants such as TLR ligands, CD40 agonists, cytokines, or activators of stimulator of IFN genes (STING) are indicated to potentiate vaccine immunogenicity and to break self-TAA tolerance. 124 Interestingly, some adjuvants can also preferentially polarize the immune cells to a Tc1-type CD8 + T cell response. Allostatine is non-toxic to immune cells, even at high dose, water soluble, stable in aqueous solutions and can be produced in large quantities at relatively low cost.
Short CD8 + peptides can bypass the need for cross-presentation by directly binding to MHC-I molecules on the surface of the APC, but they show low efficiency due to the lack of CD4 + T cell help, illustrating the importance of cross-presentation. It is well-established that CD4 + T cells and especially Th1 cells play a key role in promoting anti-tumor cellular immunity, mediated via their "helper" function. 136 Activation of CD4 + T cells is essential for optimal CD8 + T cell priming, by enhancing antigen presentation and favoring cross-priming of tumor antigen by DC 137 (see also Figure 4). CD4 + T cells also support recruitment of CD8 + T cells to the tumor site, maintenance, and expansion of a CD8 + T cell memory response. 138 Besides their "help" to CD8 + T cell responses, CD4 + T cells can mediate other anti-tumor effects such as direct killing of tumor cells, recruitment toward and activation of innate immune cells (eg, NK cells or M1 macrophages) at the tumor site, and modulation of the tumor microenvironment by anti-angiogenic effects. 139,140 CD4 + Th1 cells also promote vessel normalization leading to better intratumoral CD8 + T cell infiltration. 141 Certain DC subpopulations, especially human CD141 + DC designated as cDC1, are thought to be specialized for cross-presentation. 142 Recent studies revealed that antibody-based targeting of antigens to specific DC receptors can enhance antigen uptake and anti-tumor vaccine efficiency. 143  The range of tumor-associated antigens recognized by T cells provides a variety of potential targets for cancer immunotherapy.
In the case of TAAs targeted via cell-mediated immunity, the acti-

| Atherosclerosis vaccines
Although promising results, as mentioned earlier in this paper,

| Therapeutic vaccines targeting human immunodeficiency virus
Human

| Therapeutic vaccines against tuberculosis
The only licensed vaccine for tuberculosis (TB) today is a prophy-

| THER APEUTI C VE TERINARY VACCINE S
While veterinary therapeutic vaccines have a rather long history, relatively few approaches have resulted in commercial products, and most of them originated from human medical platforms. Initial attempts focused on oncological diseases in production animals. In 1958, an autologous vaccine was developed for bovine papillomatosis, a viral disease of cattle characterized by the presence of mucosal and cutaneous neoplastic lesions. It was synthesized from formalininactivated supernatants of affected tissue. 196 The autologous vaccination procedure reportedly induced remission of the skin lesions in diseased cattle. 196,197 However, the results were based on clinical observations, and experimental data are lacking from the literature.
Recent publications showed high recovery rates, using therapeutic schemes related to human papilloma virus vaccines, but none of them reached the market (reviewed by 198).
The causative agent of bovine papillomatosis, the bovine papilloma virus, is also involved in the pathogenesis of equine sarcoid, the most frequent skin tumor in horses. 199 Several research groups investigated the efficacy of autologous vaccines against equine sarcoids. [200][201][202] According to these publications, complete tumor regression was noted in a high percentage of affected horses.
However, the number of studied cases was limited, while other strategies, such as chemotherapy or surgical removal of the lesions, have been suggested as alternative treatments. 203 The current absence of commercially registered therapeutic tools favors the development of small private laboratories, which are specialized in the production of autologous vaccines, not only for equine sarcoids but also for other malignancies of companion animals.
In contrast to the human medical sector, the pharmaceutical in-

| FUTURE PER S PEC TIVE S
Since a majority of chronic diseases affecting humans and animals have an underlying immunological basis, therapeutic vaccine approaches can be envisaged to intervene, to eliminate or reduce disease symptoms. As a result, therapeutic vaccination has great potential to address a range of infectious and non-infectious conditions. However, there are many challenges in comparison with prophylactic vaccination including disease heterogeneity, a requirement to block or divert ongoing immune responses, and a lack of clarity on the optimal antigens. Progress in this field will depend on developing novel adjuvant approaches and appropriate tailoring of adjuvants to achieve a specific goal. Given the enormous diversity in requirements from driving potent tumor-specific T cells in the case of cancer to a dominant Treg response for autoimmune diseases, the types of adjuvant required will be diverse and it would be a major step forward if a panel of clinically acceptable adjuvants that can achieve these diverse outcomes were made available. While there are very promising advances in antigen discovery, exciting recent developments, particularly in situ vaccination for cancer, 208 offer the potential for adjuvant strategies to act as therapeutics without the challenge of identifying and incorporating specific antigens. Related to this, innate immune training presents strategies to modulate innate immunity to drive effector responses or reduce inflammation in an antigen-independent manner and strategies toward identifying the optimal adjuvants to achieve this are being evaluated. 209 In the case of cancer (and autoimmune diseases), it is likely that vaccination approaches will be used in parallel with other therapeutic strategies. In particular, for cancer vaccines, optimal combinations of therapeutic vaccines and checkpoint inhibitors are likely to be more effective than the vaccines alone. In the case of autoimmune diseases, therapeutic vaccines that reduce inflammatory T cell responses could be combined with anti-cytokine therapies to bolster anti-inflammatory responses and reduce pathology. Overall, although this is a challenging field, the future appears bright and there will be a major role for adjuvant researchers to facilitate the generation of innovative new therapeutic vaccine strategies.

ACK N OWLED G EM ENT
This article/publication is based upon work from COST Action CA16231 ENOVA (European Network of Vaccine Adjuvants), supported by COST (European Cooperation in Science and Technologywww.cost.eu).

CO N FLI C T O F I NTE R E S T
Jens Brimnes is an employee at ALK-Abello A/S, a company that manufactures and sells allergy immunotherapy (AIT) products.
Sergey Chernish is the owner of Alopharm, a company producing