Bidirectional crosstalk between therapeutic cancer vaccines and the tumor microenvironment: Beyond tumor antigens

Abstract

Immunotherapy has rejuvenated cancer therapy, especially after anti-PD-(L)1 came onto the scene. Among the many therapeutic options, therapeutic cancer vaccines are one of the most essential players. Although great progress has been made in research on tumor antigen vaccines, few phase III trials have shown clinical benefits. One of the reasons lies in obstruction from the tumor microenvironment (TME). Meanwhile, the therapeutic cancer vaccine reshapes the TME in an ambivalent way, leading to immune stimulation or immune escape. In this review, we summarize recent progress on the interaction between therapeutic cancer vaccines and the TME. With respect to vaccine resistance, innate immunosuppressive TME components and acquired resistance caused by vaccination are both involved. Understanding the underlying mechanism of this crosstalk provides insight into the treatment of cancer by directly targeting the TME or synergizing with other therapeutics.

1. Introduction

Cancer, one of the major killers, seriously threatens people's health and lives. Recent research has proven the potential capability of the human immune system to eliminate cancer [1]. The human immune system has three main functions: surveillance, defense and immune regulation. Of note, dysfunction of surveillance leads to cancer, while failure of defense and immune regulation correlates to infection and autoimmune diseases, respectively [2]. Evasion from immune surveillance has been recognized as a hallmark of cancer [3]. In turn, immunotherapy has emerged as a new approach for eradicating malignant tumors. As one of the most developed parts of immunotherapy, therapeutic cancer vaccines carry great expectations, especially after the FDA's approval of Sipuleucel-T as the first cancer vaccine for prostate cancer in 2010 and Seviprotimut-L for melanoma in 2020 [4,5]. The first cancer vaccine, an allogeneic melanoma lysate, was reported in 1998 [6]. Many attempts have focused on peptide vaccines, dendritic cell (DC) vaccines, genetic (mRNA or DNA) vaccines, viral vaccines and in situ cancer vaccines (ISVs). The advantages and disadvantages of different types of vaccines are listed in Table 1. As is widely known, most cancer vaccines focus on the activation of tumor-specific cytotoxic CD8⁺ T cells due to their key therapeutic role in mouse models [2]. Activated cytotoxic CD8⁺ T cells usually wipe out cancer cells by recognizing particular MHC class I restricted peptides, which are processed from tumor antigens and expressed on their surface. Furthermore, tumor cell lysis leads to the release of a variety of tumor antigens, which are taken up by endogenous antigen presenting cells (APCs), such as DCs. These newly released antigens will then be presented to additional T cells by APCs, resulting in a broader antitumor response [7]. Additionally, several cytokines and toll-like receptor (TLR) ligands, such as IFN-γ, IL-2 and bacterial lipoproteins, can promote in vivo presentation by endogenous APCs [8]. More recently, therapeutic cancer vaccines were developed and primarily target tumor-associated antigens (TAAs) or tumor-specific antigens (TSAs) to achieve a chronic therapeutic response by leveraging immunologic memory.

Table 1.

Comparison of cancer vaccines.

Cancer vaccine	Advantage	Disadvantage
Peptide vaccine	Be easily produced and is not oncogenic	Must be HLA class I and II restricted [1]
DC vaccine	Quickly activate T cells without considering patients’ HLA type [2]	Complexity of procedure and the maturation of DCs [1]
mRNA vaccine	Not be integrated into the host genome; not be restricted by HLA typeAllow the combination of mRNAs encoding different antigens [3]	Instability and short half-life [3]
DNA vaccine	Simple, stable, cost effective and can encode both tumor antigens and immunomodulators [4]	Not easy to break peripheral T cell tolerance; difficult to spread from cell to cell; readily be taken up or expressed by DCs [4]
Viral vaccine	Naturally immunogenic [5]	Be disturbed by patients’ neutralizing antibodies to viral vectors [5]

^[1]L. Li, S.P. Goedegebuure, W.E. Gillanders, Preclinical and clinical development of neoantigen vaccines, Ann Oncol, 28 (2017) xii11-xii17.

^[2]K.F. Bol, G. Schreibelt, K. Rabold, S.K. Wculek, J.K. Schwarze, A. Dzionek, A. Teijeira, L.E. Kandalaft, P. Romero, G. Coukos, B. Neyns, D. Sancho, I. Melero, I.J.M. de Vries, The clinical application of cancer immunotherapy based on naturally circulating dendritic cells, J Immunother Cancer, 7 (2019) 109.

^[3]L. Miao, Y. Zhang, L. Huang, mRNA vaccine for cancer immunotherapy, Mol Cancer, 20 (2021) 41.

^[4]A. Lopes, G. Vandermeulen, V. Préat, Cancer DNA vaccines: current preclinical and clinical developments and future perspectives, J Exp Clin Cancer Res, 38 (2019) 146.

^[5]E. Sasso, A.M. D'Alise, N. Zambrano, E. Scarselli, A. Folgori, A. Nicosia, New viral vectors for infectious diseases and cancer, Semin Immunol, 50 (2020) 101430.

However, there are many challenges in the development of tumor vaccines. Despite the approval of Sipuleucel-T and Seviprotimut-L, a series of large phase III cancer vaccine studies reported negative outcomes, and the other cancer vaccines are currently still in clinical phase I or II [4,[9], [10], [11], [12], [13]] (Table 2). Cancer vaccines targeting self-molecules such as TAAs have proven difficult. As reported, TAAs are poorly immunogenic and often heterogeneously expressed by genetically unstable tumor cells that undergo mutations, which cannot commonly stimulate immune responses [14]. In addition to intrinsic factors, several extrinsic factors help tumor cells gradually develop strategies to escape recognition by the immune system. As one of the key external factors, the role of the tumor microenvironment in tumor immunotherapy has attracted extensive attention.

Table 2.

Cancer vaccines in phase III trials.

Cancer vaccine	Combination	NCT	Disease	N	Recruitment status	Primary outcome measures	Study start date
ProstAtak®	RT	NCT01436968	Localized Prostate Cancer	711	Active, not recruiting	DFS	2011/9
M-Vax	No combination	NCT00477906	Metastatic Melanoma	387	Unknown	Best overall anti-tumor response & Survival	2016/7
PROSTVAC-V/F	No combination	NCT01322490	Metastatic Prostate Cancer	1297	Completed	OS	2011/11
Imprime PGG®	Cetuximab (Erbitux®)	NCT01309126	Colorectal Cancer	217	Terminated	OS	2011/4
EGF Vaccine	No combination	NCT02187367	Non-small Cell Lung Cancer	106	Terminated	OS	2015/5
BiovaxId (FNHLId1)	No combination	NCT00091676	Non-Hodgkins Lymphoma	629	Unknown	DFS	2000/1
G17DT	No combination	NCT02118077	Pancreatic Cancer	154	Completed	Survival	2001/4
CG0070	No combination	NCT01438112	Bladder Cancer	22	Terminated	CR & DCR	2014/3
LIT/inCVAX	No combination or Cyclophosphamide	NCT03202446	Advanced Breast Cancer	18	Terminated	ORR	2016/6
DCVax®-L	No combination	NCT00045968	Newly Diagnosed GBM Brain Cancer	348	Unknown	PFS	2006/12
OSE2101	No combination	NCT02654587	Advanced Non-small Cell Lung Cancer	363	Active, not recruiting	OS	2016/2
PV-10	No combination	NCT02288897	Locally Advanced Cutaneous Melanoma	20	Terminated	PFS	2015/4
Racotumomab	No combination	NCT01460472	Advanced Non-small Cell Lung Cancer	1082	Unknown	OS	2010/9
DCVAC/PCa	Chemotherapy	NCT02111577	Metastatic Castration-resistant Prostate Cancer	1182	Completed	OS	2014/5
DCVAC/PCa	Chemotherapy, Bevacizumab, PARPi	NCT03905902	Ovarian Cancer	0	Withdrawn	OS	2021/8
Toca 511	Toca FC	NCT02414165	High Grade Glioma	403	Terminated	OS	2015/11
OncoVAX	No combination	NCT02448173	Stage II Colon Cancer	550	Recruiting	DFS	2015/5

The concept of the tumor microenvironment (TME) can be traced back to the classical "seed and soil" theory proposed by Paget in 1889, placing research of the tumor-surrounding environment on the agenda [15]. The definition of the TME has been gradually improved, comprising cellular and noncellular components surrounding the tumor, such as blood and lymph vessels, immune cells, fibroblasts, signaling molecules, exosomes and the extracellular matrix (ECM). The formation of the TME is the result of the interplay between tumor cells and the body's defense system. Tumor cells release extracellular signals to recruit cells and nutrients for proliferation and migration, while the immune system dispatches immune cells to kill malignant cells, orchestrating the complexity of the TME. The characteristics of the TME include hypoxia, immunosuppression, chronic inflammation, acidosis, high interstitial fluid pressure, increased ECM stiffness and depletion of essential nutrients. The former two are hotpots of current research as important targets of immunotherapy [16,17]. Notably, therapeutic cancer vaccines activating the local immune TME can promote the induction of tumor cell death, implicating the vital role of the TME in therapeutic cancer vaccine efficacy [14]. Hence, the TME has been increasingly recognized as a valid objective for therapeutic intervention, and more light has been shed on cancer vaccines targeting the TME.

In view of the importance of the TME for vaccines, this review introduces how therapeutic cancer vaccines affect TME components, what role the TME plays in cancer vaccine resistance and promising vaccines targeting the TME (Fig. 1).

Fig. 1.

Bidirectional crosstalk between therapeutic cancer vaccines and the TME. Vaccines can be classified into five types: peptide vaccines, DC vaccines, mRNA vaccines, DNA vaccines and viral vaccines. Vaccines influence TME by activating or inhibiting immunity, while TME, in turn, plays a role in vaccine potency with dual effects. IL-12 stimulates NK cells, while GM-CSF indirectly activates T cells by facilitating DC maturation. In response to γ-radiation, tumor cells secrete exosomes, which also stimulate DC mutation. The B-cell–activating factor (BAFF) in vaccines can also lead to downstream CD4⁺ T-cell activation and a memory phenotype of T cells, which actives vaccine potency as well. In contrast, aberrant angiogenesis helps tumors escape immunity. Cancer-associated fibroblasts (CAFs) also inhibit vaccine potency. Some metabolites, such as β-GC, fatty acids and indoleamine 2,3-dioxygenase (IDO), suppress the function of immune cells, which further impairs vaccine effects. Vaccines can exert immune effects on tumor cells with the help of immune cells such as DCs and NK cells. In addition, they can induce immunogenicity, likely through VEGF-C-driven lymphangiogenesis. However, inhibitory receptors expressed on CD8⁺ T cells, such as PD-1 and NKG2A, lead to limited immunity. IFN-γ and the sympathetic nervous system are also involved.

2. The TME influences vaccine potency: Ally or barrier

The TME is involved in vaccine therapy in a highly ambivalent way. Some components in the TME promote the tumor-killing effect of tumor vaccines, while other components serve as safeguards to tumor cells by providing mechanical support or secreting different cytokines to evade treatment. This ambivalence endows cancer vaccines with inhomogeneous potency in patients. The final outcome depends on the contribution of all TME components, mainly immunosupportive and immunosuppressive components.

2.1. The immunosupportive microenvironment

2.1.1. Supportive cells and cytokines

Tumor-antagonizing immune cells primarily consist of effector T cells (including cytotoxic CD8⁺ T cells and effector CD4⁺ T cells), B cells, natural killer (NK) cells, DCs, M1-polarized macrophages and N1-polarized neutrophils. In the TME, most macrophages and neutrophils are educated by tumors to be type II immunosupressive cells; thus, cancer vaccines are mainly focused on T cells, B cells, NK cells and DCs. [18,19].

Cytotoxic CD8⁺ T lymphocytes (CTLs) are the primary immune cells capable of killing tumor cells, and they are activated by antigens presented by MHC-I molecules via cross presentation. Vaccine-specific CTL responses are positively correlated with prolonged time to progression and overall survival, further highlighting the important role of CTLs [20,21]. Therefore, most current cancer vaccines focus on the activation of tumor-specific CTLs [2]. However, the key role of effector CD4⁺ T cells has been gradually realized. They can induce and maintain the responses of antitumor CD8⁺ T and B cells and elicit long-lasting immunological memory [22], [23], [24]. AE37, a widely known HER2-targeted vaccine, is engineered to induce a strong CD4⁺ T-cell response and has achieved clinical improvements, while another two HER2-specific vaccines, E75 and GP2, induce a CD8⁺ CTL response to HER2 [25], [26], [27]. In addition, recent research has pointed out that AE-37-induced CD4⁺ T helper cells account for engagement of DCs at tumor sites, promoting cross-presenting antigens from apoptotic tumor cells and epitope spreading, thereby activating antitumor CD8⁺ T cells [28,29]. B cells can also facilitate cancer vaccines as APCs. B-cell-activating factor (BAFF) functions as a B-cell immune checkpoint because of its necessity to activate B cells, but it is rarely supplied in the TME. Application of whole-cell vaccines with BAFF upregulates B-cell costimulatory molecules (CD40 and ICOSL), leading to downstream CD4⁺ T-cell activation and a memory phenotype of T cells in vaccine-draining lymph nodes [30]. In a B16F10 melanoma mouse model, the BAFF vaccine augments antitumor efficacy with a more remarkable delay in tumor growth [31]. NK cells are the primary immune cell population that eliminates malignant cells autonomously without prior sensitization or recognition of specific tumor antigens [32]. Clinically, activation of NK cells has been proven essential for better clinical outcomes in a variety of cancer vaccines [33], [34], [35]. They can promote antigenic presentation by inducing the maturation of DCs, which are one of the major APCs [36]. Notably, DC-based vaccines are the mainstream vaccines due to their strong immunoactivating capacity. NK-cell-dependent lysis of immature DCs prevents resistance to DC vaccines, indicating that NK cells in combination with DCs are important for vaccine-induced antitumor processes [37]. M1-polarized macrophages produce proinflammatory cytokines and reactive oxygen/nitrogen species to destroy malignant cells. N1-polarized neutrophils release cytotoxic granules and cytokines/chemokines to destroy malignant cells, secreting cytokines and chemokines to recruit other cells with antitumor activity [38].

In addition to supportive immune cells, it cannot be ignored that many cytokines are also involved in immune stimulation by tumor vaccination. Interestingly, most of them are secreted by the supportive cells mentioned above and are also essential for the antitumor effects of several supportive immune cells. For instance, IFN-γ produced by CD8⁺ T cells is indispensable for tumor regression [39]. Interleukin (IL)-12 is an important cytokine involved in a variety of antitumor pathways. First, it is a potent inducer of the activation of NK cells and CD8⁺ T cells, both of which are the main force of tumor elimination. In addition, it can exert an antitumor effect in an indirect way. It can induce the differentiation of naïve CD4⁺ T cells to Th1 cells. Th1 cells can directly activate CTLs by secreting IL-2 and then enhance the antitumor response [40]. In addition, Th1 cells can also indirectly activate CTLs by secreting IFN-γ, contributing to enhanced DC-dependent antigen presentation to CTLs. Moreover, paracrine release of IL-12 stimulates IFN-γ production, enhancing vaccination potency with a novel peptide-based cancer vaccine in an OT-1 TCR transgenic mouse model [41]. In addition to IL-2 and IFN-γ, IL-12 also plays an important role in inducing potent anticancer effects by synergizing with several other cytokines, such as IL-15, granulocyte-macrophage colony-stimulating factor (GM-CSF), and tumor necrosis factor (TNF)-α [42]. Among these cytokines, GM-CSF has been broadly applied as a vaccine adjuvant to increase the T-cell-dependent immune response. GM-CSF was initially discovered to induce the development of both granulocytes and monocytes and macrophages but then be broadly incorporated into vaccines to stimulate potent antitumor responses, likely by promoting the differentiation of cross-presenting DCs [43]. On the one hand, GM-CSF functions to stimulate adaptive immunity by enhancing the local recruitment of DCs to the vaccine site and subsequently increasing antigen presentation, thereby indirectly activating antitumor CD8⁺ T cells [44]. On the other hand, some research revealed that GM-CSF-secreting breast cancer vaccines could elicit CD4⁺ T-cell immunity because of an upregulated immune signature [45]. Apart from GM-CSF, polyactin A can also efficiently induce mature DCs from PBMCs with IL-4 and TNF-α for a peptide vaccine [46].

Taken together, the induction of immunosupportive cells and cytokines may be the critical task for the success of cancer vaccines.

2.1.2. Exosomes are important in vaccination

Exosomes from immune cells or tumor cells are highly related to vaccine effectivity. Exosomes are nanometric membrane vesicles that can be released by almost all cell types. They can facilitate the proliferation, activation and apoptosis of T cells and are therefore conducive to vaccine immunity. Macrophages are categorized into M1 and M2 cells, supporting and suppressing the Th1 immune response, respectively [47], [48], [49]. Different from M2 macrophages, exosomes derived from the M1 type have the potential to be a new class of immune adjuvants because they enhance the activity of Trp2 vaccines toward melanoma by conveying a stronger antigen-specific cytotoxic T-cell response [50]. Additionally, tumor cells can secrete exosomes under γ-radiation stimuli. Of note, exosomes from γ-irradiated melanoma cancer cells, named G-exos, can induce more mature DCs and IFN-γ-producing CD8⁺ T cells, ultimately resulting in enhanced vaccine efficacy to inhibit tumor growth [51]. Major histocompatibility complex (MHC) molecules are involved in the process of antigen presentation. Exosomes from an MHC class II transactivator (CIITA) gene DC vaccine significantly increased the expression of IFN-γ-producing CD4⁺ T cells while upregulating TNF-α and IL-12 but inhibiting negative IL-10 secretion [52]. Overall, exosomes are essential for vaccines to exert antitumor effects, suggesting their promising roles in vaccine design and development.

2.2. The immunosuppressive microenvironment: innate resistance to cancer vaccines

2.2.1. Suppressive cells and cytokines

Immunosuppressive cells in the TME mainly consist of myeloid-derived suppressor cells (MDSCs), tumor-associated macrophages (TAMs), regulatory T cells (Tregs) and cancer-associated fibroblasts (CAFs), facilitating peripheral immune tolerance or immune evasion [53]. Published work proved the association between Tregs/MDSCs and impaired vaccine efficacy, indicating that immunosuppressive cell levels may represent a poor prognosis after cancer vaccination and serve to identify patients who probably do not benefit from vaccine therapy [54].

As the dominant cell type in the solid TME, CAFs have been considered crucial in immune suppression. Previous studies have shown that CAF infiltration indicates unfavorable potency of vaccines. One possible reason is that CAFs construct a compact ECM to separate tumor cells from T effectors and chemotherapy drugs in contrast to normal fibroblasts, helping them avoid immune attack [55]. Unlike melanoma or lung squamous cell carcinoma, pancreatic cancer is quite resistant to vaccination therapy, which is thought to be partially due to the dense fibrotic tumor microenvironment [56]. CAFs suppress tumor responses to anticancer vaccination by promoting CD8⁺ T-cell exclusion from tumors, which is mechanically associated with upregulation of cytotoxic-T-lymphocyte-antigen-4 (CTLA-4) in CD8⁺ T cells by NOX4 inhibition [57]. CAFs can also suppress CTLs in an indirect way. They can recruit monocytes to the TME and then support their differentiation into type 2 TAMs, which inhibit the activity of CTLs [58].

Apart from immune cells, immunosuppressive factors also participate in immune tolerance and evasion. Usually, most suppressive factors are produced by cancer cells and some immunosuppressive cells, such as Tregs and TAMs. Common immunosuppressive factors mainly include vascular endothelial growth factor (VEGF), transforming growth factor (TGF)-β, indoleamine 2,3-dioxygenase (IDO), prostaglandin E2 (PGE2), IL-10, and IL-6.

Similar to supportive ones, they also function with immune cells to tamp down the chemotactic movement and function of cytotoxic CD8⁺ T cells, suggesting poor clinical results of cancer vaccines. One example is DCs and relevant cytokines. Despite initiation of the antitumor immune response performed by mature DCs, immature or regulatory DCs impede the immune response. Immature DCs are involved in the induction of T-cell anergy, generation of Tregs, and promotion of alloantigen-specific tolerance [59], while regulatory DCs induced by tumor cells inhibit the T-cell response with the help of high expression of immunosuppressive factors such as IL-10, nitric oxide (NO), VEGF and arginase I [60]. Thus, tumor cells can educate DCs to lose antitumor function and even differentiate into a regulatory DC subset with the help of cytokines, promoting tumor immune escape.

Interestingly, several cytokines have dual roles. For instance, the Nlrp3 inflammasome is a proteolysis complex that has contrasting roles in tumorigenesis, demonstrating both detrimental and beneficial effects, with detrimental effects being observed in most cases [61]. In the presence of IL-12, activation of the Nlrp3 inflammasome diminishes antitumor immunity and limits the efficacy of the DC vaccine by facilitating the migration of MDSCs to the tumor site [62]. These components in the TME mediate immune escape from vaccination and promote tumor progression. Except for immune cells, immunosuppressive factors, such as vascular endothelial growth factor (VEGF), transforming growth factor (TGF)-β, indoleamine 2,3-dioxygenase (IDO), prostaglandin E2 (PGE2), IL-10, and IL-6, are frequently produced by cancer cells, Tregs or macrophages. By being involved in immune escape, these cells and factors can tamp down the chemotactic movement and function of cytotoxic CD8⁺ T cells and the clinical results of cancer vaccines.

Although granulocyte colony-stimulating factor (G-CSF) is traditionally considered an immunosupportive cytokine, its reverse function related to MDSC expansion has been recently discussed. Breast cancer cells that express high levels of G-CSF tend to abrogate the efficacy of breast cancer autologous tumor cell vaccines [63]. In addition, G-CSF inhibition enhanced the immunogenicity of a 4T1-based vaccine [64]. Given that G-CSF is routinely administered to prevent neutropenia in breast cancer patients with chemotherapy, cautious consideration is warranted.

Most cancer vaccines are tested in advanced cancers, where the immunosuppressive TME is dominant. Therefore, the profound understanding and inhibition of suppressive cells and cytokines needs more attention. Cancer vaccines with the ability to thwart immunosuppression appear to be a prerequisite for the next generation of cancer vaccines.

2.2.2. The metabolic microenvironment

Modified metabolism is a hallmark of cancer [65]. The metabolic TME mainly includes decreased concentrations of nutrients and oxygen and increased levels of harmful metabolites. These factors may contribute to changes in the immune landscape, finally leading to the formation of an “immunologically cold” cancer that is insensitive to vaccination.

Tumor cells show deregulated metabolism, leading to a hypoxic TME. Researchers have confirmed the role of the hypoxic TME in the maintenance of tumors and resistance to therapies by regulating the cancer stem cell population [66,67]. In addition, hypoxia also promotes tumor immune escape by reducing T-cell infiltration and activating several oncogenic pathways, which significantly lowers vaccine potency [66]. Likewise, recruitment of inhibitory immune cells is also an effective strategy for tumors to escape immune surveillance. For instance, hypoxia can directly induce TAM infiltration around tumors and then promote their polarization toward the immunosuppressive M2 phenotype [68]. Luckily, researchers have developed some potential countermeasures. It was found that the oxygen-generating nanoparticle NanoMnSor can efficiently modulate the hypoxic TME and increase the infiltration of type 1 immunosupportive TAMs and cytotoxic CD8⁺ T cells in tumors, augmenting the efficacy of immunotherapy, including whole-cell cancer vaccines [38,69]. Therefore, hypoxia modulation would be beneficial to enhance the immunogenicity of cancer vaccines and provide long-term immune memory.

In addition to the hypoxic environment, carbohydrates, lipids and amino acids must always be the protagonists in regard to metabolism. β-GC is a metabolic intermediate in the anabolic and catabolic pathways of complex glycosphingolipids. β-GC synthase deficiency was shown to impair NKT-cell activation [70]. NKT cells belong to a subset of T cells that can secrete Th1 and Th2 cytokines to influence antitumor activity. In clinical applications, β-GC serves as an NKT-cell stimulator to enhance the anti-HCC effects of HBsAg gene-modified DC vaccines [71].

Another metabolite consistently upregulated across multiple tumor types is prostaglandins, which are physiologically active unsaturated fatty acids and have very potent immunosuppressive properties [72]. Fatty acid metabolism influences the function of immune cells. Of note, cyclooxygenase (COX), a key enzyme catalyzing the production of prostaglandins, including PGE₂, has been found to profoundly inhibit inflammatory cytokine IFN-γ and IL-2 production while enhancing anti-inflammatory cytokine IL-4 production. Furthermore, PGE₂ has been shown to prevent DC maturation and promote CD4⁺/CD25⁺ Treg activities, which strongly invades the antitumor immune response [73]. Accordingly, the pancreatic TME accumulates very-long-chain fatty acids during cancer progression, leading to the mitochondrial defect of intrapancreatic CD8⁺ T cells and finally inducing T-cell dysfunction [74]. Accordingly, targeting fatty acid metabolism may be a potential way to enhance the efficacy of treatment. For example, PPAR agonists can increase the efficacy of a melanoma cancer vaccine by increasing fatty acid metabolism and providing T cells access to glucose in the TME [75]. Interestingly, levels of plasma membrane cholesterol also correlate with the effector function of CD8⁺ T cells. A Kras-peptide-based vaccine combined with the cholesterol modulator avasimibe potentiates the anti-Kras-driven lung tumor effect by reducing the numbers of Treg cells and promoting the expansion of effector T cells [76].

IDO, an enzyme that limits the depletion of tryptophan, is found to be key in regulating immune tolerance by causing T-cell suppression [77], suggesting the dual roles of tryptophan metabolism in tumor progression. In fact, IDO-positive DCs are related to the generation and maintenance of peripheral tolerance via the induction of Tregs [78]. IDO inhibition is considered to be crucial in clearing tumors. Via suppression of IDO expression, glycogen synthase kinase-3β (GSK-3β) inhibition can enhance DC-based cancer vaccine potency [79]. Overall, the metabolic TME has the potential to be an actionable target to enhance the efficacy of DC‑based cancer vaccines.

2.2.3. Angiogenesis

Identical to the metabolic microenvironment, abnormal angiogenesis is also a hallmark of cancer [65]. Tumor-associated vasculature provides essential nutrients and oxygen for solid tumors, which are always dysfunctional due to their heterogeneity and tortuousness. Tumor growth is generally limited to 1–2 mm³ in the absence of a blood supply [80]. Vascular abnormalities facilitate inefficient blood perfusion, hypoxia, and finally an immunosuppressive TME, characterized by increased infiltration of Tregs, MDSCs, and M2- TAMs and reduced infiltration of CTLs [81]. Antiangiogenic therapy could increase tumor infiltrating lymphocytes (TILs) around tumors [82]. Vascular endothelial cells express angiogenic molecules, including VEGF, which impairs DC maturation and antitumor immune activity [83]. Correspondingly, VEGF expression is negatively related to vaccination potency. For instance, an antimicrobial peptide called GE33 exhibits high antitumor specificity by decreasing the expression of VEGF [84].

3. Cancer vaccines reshape the TME

In section 2, we described the major components of the TME that increased or decreased the efficacy of vaccines. We also briefly mentioned several potential strategies by targeting the TME. Actually, the TME is a dynamically changing setting. After vaccination, a few immune components are recruited and activated, while others are excluded and inhibited, leading to a change in the TME landscape. The consequence of this change is immune stimulation or escape. Here, we will further introduce how vaccines reshape the TME from positive and negative aspects at length.

3.1. Vaccination activates immunity: convert the cold TME to the hot TME

Primary resistance to tumor vaccines is prominently associated with low tumor-infiltrating immune cells, so-called cold tumors. Great hopes are placed on turning cold tumors into T-cell-infiltrated hot tumors by augmenting immunostimulation and abrogating immunosuppression at the tumor site [85]. Vaccination is capable of reversing immunosuppression in the TME, which greatly enhances therapeutic potency [86] (Fig. 2). For example, a chaperone-rich tumor-cell lysate (CRCL) vaccine partly reverts the tumor-induced suppression of APCs and delays the progression of pre-established tumors [87]. Similarly, a DNA vaccine encoding both the NKG2D ligand H60 and TAA survivin can induce a more markable tumor reduction and maintain a longer T-cell memory by increasing crosstalk between lymphocytes [88]. Notably, NKG2D is expressed on NK cells, activated T cells and macrophages. It affects both innate and adaptive immunity, indicating its promising role in breaking immune tolerance [89]. Certainly, vaccines can alter the TME in several other ways, which will be described below.

Fig. 2.

Vaccination activates immunity via the TME. Cancer vaccines mobilize the TME to attack malignant cells. DC is the prominent cell vaccine. NKG2D-loaded DCs can activate NK cells and then induce marked tumor reduction. GM-CSF and IL-10 are common stimulators of DCs for better antitumor immunity. In addition, DC immunogenicity is augmented by TLR agonists such as Poly (I:C). Downstream, DC vaccines regulate the NK subpopulation and enhance NK activation. Meanwhile, metabolites linked to phosphatidylserine (PS) can inhibit DC maturation, thus contributing to TGF-β and IL-10 secretion and poor immunogenicity. Vaccines loaded with annexin 5A or 2aG4 mask PS, showing satisfactory potency. Interestingly, VEGF-C-driven tumor lymphangiogenesis promotes an immunosupportive TME, which means that VEGF-C can be a clinical target.

3.1.1. Augmenting DC immunogenicity to propel antigen presentation

DCs are the most efficient APCs of the innate immune system [90]. Tumor-induced DC immunosuppression remains the major roadblock for developing effective vaccines. To further liberate DC vaccine potency, some attempts have been focused on promoting DC maturation and costimulation by selectively increasing the exposure/release of particular pathogen/damage-associated molecular patterns (PAMPs/DAMPs), potentially giving rise to a next generation of potent Th1-driven DC vaccines [91]. Upon sensing threats, immune cells respond to PAMPs/DAMPs through pattern recognition receptors (PRRs), such as Toll-like receptors (TLRs), and secrete cytokines and chemokines to potentiate the innate immune response. Interestingly, in addition to immune cells, tumor cells also have the potential to enhance the antitumor response. It has been reported that tumor cells undergoing immunogenic cell death (ICD) exhibit excellent immunostimulatory capacity owing to the spatiotemporally defined emission of a series of critical DAMPs that can trigger the immune system [92]. Therefore, TLR agonists and ICD inducers seem to be exploited to enhance adaptive responses by potentiating DC maturation and immunogenicity during vaccination.

3.1.1.1. TLRs

Activation of the Toll pathway results in the production of proinflammatory cytokines and chemokines and DC maturation, which in turn promotes CTL responses linking innate and adaptive immunity. TLRs can be classified as surface TLRs and intracellular TLRs. In different manners, DC vaccines modified with these two types of TLRs could be more powerful.

Surface TLRs include TLR1, TLR2 and TLR4–6. They sense bacterial, fungal, and protozoal products on the cell surface. It has been reported that they are associated with the production of many cytokines and chemokines, such as IL-6, IL-12, TNF and IL-8, contributing to the antitumor immune response [93]. PAMPs recognized by surface TLRs induce more mature DCs (mDCs) to stimulate specific anticancer immune responses. These well-studied PAMPs include lipopolysaccharide (LPS) and GM-CSF (detected by TLR4) [94], [95], [96], bacterial lipoproteins and lipoteichoic acids (detected by TLR2), flagellin (TLR5 agonist) [97] and Pam3CSK4 (TLR2 agonist) [98]. SRA/CD204 is a PRR that acts as a negative immune regulator via downregulation of TLR4. DC-mediated silencing of SRA/CD204 greatly improved the potency of DC vaccines, leading to better control of tumor progression [99]. TriMix is a mixture of three mRNAs encoding CD40 ligands, TLR4 and CD70, which have already been confirmed to modulate DC functionality and facilitate CD8⁺ T-cell maturation and activation [100,101]. Vaccination with DCs electroporated with TAA and TriMix mRNA led to the induction of durable antitumor responses in melanoma patients [102,103].

In contrast to surface TLRs, intracellular TLRs, including TLR3 and TLR7–9, are mainly responsible for sensing viral nucleic acids. They reside in intracellular organelles and produce type I IFNs [93]. Specifically, intracellular TLRs usually detect the unmethylated CpG DNA of bacteria and viruses (detected by TLR9) [104], double-stranded RNA (detected by TLR3), and single-stranded viral RNA (detected by TLR7/8) [105,106]. The commonly used agonist of endosome TLRs, especially TLR3, is poly I:C [107], [108], [109]. Poly(I:C) is a prototypical immunostimulatory dsRNA analog that potently stimulates innate pattern recognition of receptors for viral RNA in macrophages and DCs. Poly(I:C) leads to the induction of type I IFNs and upregulated PD-L1 expression on the surface of mouse CD8α⁺ DCs [107,108]. The FixVac DC vaccine stimulates the secretion of TLR7 and IFN-α to promote T-cell priming and maturation, which is maintained by continued vaccination for more than one year [110]. In addition, a pilot trial showed that poly I:C intratumor injection was well tolerated in patients with solid tumors, as evident by the patients achieving clinical benefit. Further investigation is currently being explored in a multicenter phase II clinical trial (NCT02423863) [111].

3.1.1.2. ICD

A few ICD inducers have been tested for the generation of immunogenic tumor cell cargo. They are then loaded in DC vaccines to form so-called ICD-based DC vaccines, including photodynamic therapy (PDT), shikonin (SK) and radiotherapy [112]. First, hypericin-PDT induces ICD through ROS-based ER stress and triggers the active emission of crucial ICD-associated DAMPs [113]. DAMPs are a group of natural endogenous adjuvants within cells. When danger signals are triggered, DAMPs will be stimulatory to the adaptive immune system. Clinically, hyp-PDT-induced ICD-based DC vaccines synergized with temozolomide chemotherapy have a high potency that clinically increases the survival of HGG-bearing mice by 300%, resulting in 50% long-term survivors [114]. Second, SK stimulates DAMPs (HSP70) and enhances ICD activity, augmenting Th1-cell proliferation. An SK tumor cell lysate (TCL)-pulsed DC vaccine efficaciously retards tumor growth and prolongs the survival of test mice [115,116]. Interestingly, radiotherapy induces a delay in tumor growth in a nonirradiated lesion, the so-called abscopal effect, by ICD-induced T-cell activation [117]. The combination of conventional treatment and DC vaccine treatment with tumor cells killed by 100 Gy radiation significantly prolongs the median overall survival (by more than 15 months) compared to a control group receiving solely conventional treatment in glioblastoma patients [118].

Regulation of DC-associated pathways or genes is also a promising approach. Compared to normal DCs, tumor-associated DCs are less mature and exhibit poor responsiveness to TLR stimulation, which is related to STAT3 hyperactivity. Therefore, the STAT3-siRNA-loaded polypeptide vaccine effectively abrogates immunosuppression in the TME by activating mature DC functionality [119,120]. Similarly, a synthetic triterpenoid, CDDO-Im, can inhibit STAT3 phosphorylation, directly leading to STAT3 transcriptional inactivation. A vaccine loaded with CDDO-Im shifts a protumor Th2 to an antitumor Th1 immune cytokine microenvironment, amplifying the antitumor functions of immune cells, including CD8⁺ T cells and macrophages [121].

Overall, DCs are effective carriers of vaccines with a natural capacity to stimulate the immune TME as TLR agonists and ICD inducers. This capacity endows DC vaccines with advantages for future applications.

3.1.2. DC vaccines upregulate NK-cell cytotoxicity

DCs can regulate NK cells and induce enhanced type 1 immunostimulatory cytokine secretion [122]. As expected, several studies have reported that DC vaccines improve the immune microenvironment by promoting NK function and converting the NK subpopulation [123]. Different kinds of DCs have different effects on NK cells. Researchers found that DCs armed with membrane-bound IL-15 promote NK-cell cytotoxicity, while IL-14⁺ DCs do not [124]. Interestingly, a similar conclusion that DCs stimulate NK activation was presented in another study, further demonstrating that this key mechanism is dependent on the cooperative activity of plasma membrane-bound TNF and IL-15 [125]. Apart from TNF and IL-15, IFN-α is also involved in antitumor processes related to DCs. Evidence from several studies proved that IFN-DCs (IFN-α-induced DCs) in follicular lymphoma patients led to the activation of NK cells and therefore enhanced the cytotoxic effects of autologous lymphoma cells. Of note, these responses indeed help improve the potency of DC cancer vaccines for follicular lymphoma patients, which was also supported by another study in DC vaccines [126,127]. Human NK cells in peripheral blood are divided into CD56^bright and CD56^dim cells, which are capable of poor and potent cytotoxic ability, respectively [127]. However, after AdV. DC injection, NK cells are biased to convert to the CD56^dim subset with potent cytotoxicity, positively impacting the clinical outcome of melanoma patients [128]. Taken together, DCs play a key role in the antitumor immune response partially by enhancing NK-cell cytotoxicity, which is one of the vital ways DC vaccines exert their efficacy.

3.1.3. Metabolism is tightly linked to the enhanced immune response

The metabolic profile of the TME is characterized by dynamic gradients of oxygen pressure, glycolysis, extracellular acidosis, an accumulation of lactate and adenosine, and depletion of essential nutrients, and most of them are closely related to the formation of the ‘cold tumor’ [17]. Comfortingly, vaccination can reverse such a negative environment and in turn stimulate CD8⁺ T cells to a degree, thereby converting the cold TME to a hot TME [129]. For example, the tumor vaccine vector TA-Met@MS promotes the transition of glycolysis to fatty acid oxidation. This photothermal metabolism regulation enhances T-cell survival and facilitates the differentiation of memory CD8⁺ T cells, thereby in turn leading to longer vaccine activity [130].

Phosphatidylserine (PS) is produced in response to oxidative stress caused by hypoxia, acidity, and metabolites accumulated during conventional chemotherapy or radiotherapy [131]. As an anionic phospholipid normally residing in the inside leaflet of the plasma membrane, it is usually exposed on cancer cells, vascular cells and stromal cells. Specifically, PS is an important player in mediating sustained immunosuppression primarily by inhibiting the maturation of DCs and stimulating DCs to secrete TGF-β and IL-10 [132]. Notably, TGF-β and IL-10 both contribute to the inhibition of antigen presentation, which is the most important function of DCs in the antitumor immune response. However, cancer cell-based vaccines are often created by killing tumor cells with high-dose irradiation or chemicals, which commonly causes a strong exposure of PS and then contributes to poor outcome of vaccines due to the poor immunogenicity  [133]. Therefore, blockade of PS is a promising strategy for enhancing the efficacy of cancer cell-based vaccines. It has been reported that masking PS with annexin 5A or 2aG4 in cancer cell vaccines reverses PS-mediated immunosuppression and dramatically enhances the induction of antitumor immune responses [134], [135], [136]. As reported in another study, IL-2 enhances 2aG4 function, indicating that the addition of IL-2 may further enhance vaccine potency [137]. It can be concluded that metabolic regulation may present valuable insight into better vaccine design.

3.1.4. Tumor lymphangion promotes an immunosupportive TME via VEGF-C

Metastatic spread depends on lymphangiogenesis. Traditionally, tumor-associated lymphangiogenesis is considered a negative factor in antitumor therapy. Nonetheless, blocking lymphangiogenesis disrupts the recruitment of naïve T cells and subsequent antitumor immunity [138]. Likewise, lymphangion promotes a more immune-infiltrated TME through the VEGF-C/VEGFR3 signaling axis [139]. In melanoma, a lymphangiogenic vaccine composed of lethally irradiated VEGF-C–overexpressing melanoma cells and topical immune adjuvants (VEGFC vax) induces VEGF-C-driven lymphangiogenesis, which boosts vaccine-induced immunogenicity [140]. These discoveries reveal that tumor-associated immunity is critically based on lymphatic vessel remodeling and drainage, which can be induced by therapeutic cancer vaccines.

3.2. Vaccination inhibits immunity: Acquired resistance to cancer vaccines

Each coin has two sides. Although cancer vaccine-induced conversion of the cold TME to a hot TME further enhances the therapeutic efficacy, vaccination also inhibits the antitumor response in some cases, which will be described in this section. Currently, it is appreciated that the immunosuppressive TME could be courted by vaccination and substantially compromises the efficacy of cancer vaccines [141]. Tumor cells can develop strategies to escape T-cell recognition boosted by vaccines, finally resulting in resistance to treatment, including vaccines (Fig. 3). Changes in the immune cell reporter and cytokine/chemokine milieu within the TME are archcriminals.

Fig. 3.

Vaccination-induced acquired resistance. CD8⁺ T cells play a dominant role in vaccine immunity. However, vaccines sometimes impair CD8⁺ T-cell function, causing immune escape and promoting tumor growth. TEGVAX stimulates TILs to secrete IFN-γ and thereby upregulates IFN-γ-dependent PD-L1 on tumor cells. mRNA vaccines can also elicit a type I IFN response. RNA nanoparticles increase PD-L1 upon activation of myeloid APCs. GVAX upregulates PD-1 on exhausted T cells and results in resistance. Vaccines can also induce overexpression of NKG2A on tumor-infiltrating CD8⁺ T cells and Qa-1b on tumor cells, leading to tumor relapse. DC vaccines induce β-AR in the sympathetic nervous system, which contributes to the inhibition of CD8⁺ T cells.

3.2.1. Upregulated PD-1/PD-L1 is correlated with suppressive cells

The binding of programmed cell death protein-1 (PD-1) and its ligand, programmed death-ligand 1 (PD-L1), mediates immune ignorance for malignant cells and is therefore closely linked to immune refractoriness [142]. PD-1/PD-L1 upregulation by vaccination leads to two impacts: protection of normal cells from immune damage and disorder of immune attack toward malignant cells, the latter so-called resistance. Taking PD-1 as an example, high PD-1 expression correlates with increased tumor-infiltrating Treg cells and reduced effector T cells, which finally leads to the immune escape of tumor cells. As expected, the GM-CSF-secreting pancreatic cancer vaccine GVAX has been found to upregulate PD-1⁺ exhausted T cells, resulting in therapeutic resistance [143]. To break this predicament, researchers and clinicians began to set their sights on the combination of anti-PD-(L)1 and corresponding vaccines, which potentially reversed the suppressive TME induced by vaccines. It has been reported that anti-PD-1 combined with vaccines can improve the survival of mouse tumor models [144]. Analogous results were also observed in preclinical models treated with the combination of GVAX and anti-PD-1 antibody therapy (GVAX/αPD-1) [145]. Similar to GVAX, TEGVAX can also induce PD-L1 expression on tumor cells in an IFN-γ-dependent manner by stimulating TILs to secrete IFN-γ, ultimately causing incomplete tumor regression. Complete elimination was achieved with administration of a PD-1 inhibitor as well [146]. Additionally, RNA nanoparticles can also increase PD-L1 expression upon activation of myeloid APCs [147]. As a countermeasure, immune checkpoint blockades (ICBs) have been used to elicit potent systemic and intratumor immunity, even in poorly immunogenic malignancies, increasing the potency of vaccines [148]. Altogether, these studies have confirmed the significant effect of PD-1/PD-L1 upregulation by suppressive cells in acquired vaccine resistance. Anti-PD-1/PD-L1 therapy is hopeful for reversion of this negative outcome.

3.2.2. IFN response is elicited by vaccination

IFNs play an ambivalent role in immunotherapy: IFNs can either elicit or block T-cell activation, proliferation, differentiation, survival and apoptosis [149]. In vivo mRNA vaccination elicits a potent type I IFN response, regardless of carrier or antigen, and interferes with cytolytic T-cell responses, finally resulting in poor outcomes [150]. Coadministration of type I IFN blockades greatly amplifies therapeutic antitumor efficacy [151]. The answer to this conundrum might reside at the timing and duration of type I IFN signaling [149].

3.2.3. Upregulation of receptors on CD8⁺ T cells contributes to resistance

NKG2A, a novel immune checkpoint, is generally expressed by NK cells but is selectively expressed by CD8⁺ T cells in the TME [152]. Therapeutic vaccination dynamically induces overexpression of NKG2A on tumor-infiltrating CD8⁺ T cells and Qa-1b on tumor cells, leading to tumor relapse. Of note, NK cells are irrelevant to this axis. The NKG2A inhibitor monalizumab can reverse this acquired resistance induced by therapeutic vaccines [153].

In cancer patients, the sympathetic nervous system, specifically adrenergic receptor (AR) signaling, may be activated by psychological strain associated with the burden and management of a life-threatening disease. β₂-AR is the most highly expressed AR subtype on immune cells, especially CD8⁺ T cells [154]. It has also been reported that DC-based cancer vaccines can induce β-AR signal conduction, which contributes to the inhibition of CD8⁺ T cells. However, propranolol, a nonselective β-blocker, strongly improved the efficacy of an antitumor STxBE7 vaccine by enhancing the quantity of TILs [155].

Collectively, cancer vaccines can induce immune escape by upregulating immune checkpoint molecules. It is necessary to reverse this change to guarantee the effect of the cancer vaccine.

4. Targeting the TME: Enhancing immune effects

As mentioned previously, the TME has a strong impact on the potency of therapeutic cancer vaccines. Correspondingly, vaccines can also reshape the TME, ultimately affecting their efficacy. In addition, it is common for TAA/TSA vaccines to incompletely eliminate malignant tumors. Therefore, nonspecific immune stimulation by a TME-targeting strategy has provided a new chance to beleaguer tumor cells and indirectly eradicate tumors, which has been briefly touched on in the previous sections (Fig. 4, Table 3). Here, we will explain both monotherapy and combination therapy in detail.

Fig. 4.

Cancer vaccine-based therapy. The TME should be considered in cancer vaccine therapy. One way is by directly targeting TME components, such as the vasculature, CAFs and some cytokines. The other and more important method is combination with other approaches to stimulate the immune response. For anti-vasculature therapy, some endothelial cell receptors are considered promising targets, such as VEGF, FGF-2 and TEM1. With respect to anti-CAF therapy, FAP and TGF-β are considered with caution. Additionally, cytokines are vital to cancer vaccines. IFN-γ enhances DC vaccine potency; GM-CSF can arm tumor cell vaccines; and other small molecules targeting IDO1 or interleukins are also promising candidates.

Table 3.

Vaccines targeting TME as monotherapies.

Cancer vaccine	Phase	NCT	Disease	N	Recruitment status	Primary outcome measures	Study start date
Allogeneic GM-CSF secreting cellular vaccine	Phase I	NCT00136422	Advanced Myelodysplasia or Acute Myelogenous Leukemia	30	Completed	Feasibility	2000/1
Allogeneic GM-CSF secreting cellular vaccine	Phase I	NCT00809588	Melanoma	70	Active, not recruiting	To determine the doses	2003/10
Allogenic GM-CSF transfected tumor cell vaccine	Phase II	NCT00389610	Pancreatic Cancer	56	Active, not recruiting	Safety	2006/9
AlloStim	Phase II	NCT02380443	Metastatic Colorectal Cancer	12	Completed	Safety, anti-tumor effect	2016/9
Autologous, Lethally Irradiated Breast Cancer Cells	Phase I	NCT00317603	Breast Cancer	15	Completed	To determine the minimum doses	2006/1
Autologous, Lethally Irradiated Breast Cancer Cells	Phase I	NCT00880464	Breast Cancer	8	Completed	To determine the minimum doses	2006/1
Bystander-Based Autologous Tumor Cell Vaccine	Phase II	NCT00101166	Melanoma (Skin)	43	Completed	Number of participants with PR	2004/10
CD40 Ligand & IL-2 Gene Modified Tumor Vaccine	Phase I	NCT00058799	Leukemia	11	Completed	Safety	1999/6
CD40L expressing and IL-2 secreting autologous B-CLL cells	Phase I	NCT00058786	Chronic Lymphocytic B-Leukemia	9	Completed	Safety	2002/12
CD40 Ligand & IL-2-expressing Tumor Cells Vaccine	Phase I	NCT00458679	Chronic Lymphocytic Leukemia	6	Completed	Adverse events	2006/12
CD40L expressing and IL-2 secreting B-CLL cells (B-CLL Vaccine)	Phase I	NCT00078520	Leukemia	9	Completed	Safety	2003/1
E39 peptide (100mcg)/GM-CSF Vaccine	Phase I|Phase II	NCT01580696	Ovarian Cancer, Endometrial Cancer	51	Completed	Safety, Local/Systemic toxicity	2012/4
GM. CD40L.CCL21 Vaccinations	Phase I|Phase II	NCT01433172	Lung Cancer, Adenocarcinoma	73	Completed	Phase I: Recommend Phase II Dose; Phase II: PFS	2012/3
GM-CSF Vaccine	Phase II	NCT00524277	Breast Cancer	456	Completed	Disease recurrence	2007/1
GM-K562 Vaccine	Phase I	NCT00442130	Chronic Lymphocytic Leukemia	27	Completed	Safety, toxicity	2007/2
GM-K562 Vaccination	Phase I	NCT00694330	Glioma	11	Completed	Feasibility, safety, biological activity	2008/6
GM-K562/leukemia Cell Vaccine	Phase I	NCT00809250	Acute Myeloid Leukemia, Advanced Myelodysplastic Syndrome	33	Completed	Safety	2008/11
GVAX	Phase I	NCT00258687	Clear Cell Sarcoma, Pediatric Renal Cell Carcinoma, Alveolar Soft Part Sarcoma, Melanoma	12	Completed	Safety, feasibility	2005/1
GVAX	Phase I	NCT01952730	Colorectal Cancer	15	Recruiting	Number of patients who fail to receive the first six scheduled vaccinations because of toxicity	2013/7
GVAX	Phase II	NCT01773395	Myelodysplastic SyndromeAcute, Myeloid LeukemiaChronic, Myelomonocytic Leukemia	123	Active, not recruiting	PFS	2013/1
HLA-A*2402 or A*0201 restricted peptides	Phase I|Phase II	NCT01949688	Solid Tumors	26	Completed	Safety,Clinical efficacy (OS)	2010/6
HLA-A2 and HLA-A3-Binding Peptides From FGF-5	Phase II	NCT00089778	Kidney Cancer	11	Terminated	Best response, Adverse events, Immunologic response	2004/9
hVEGF26-104/RFASE	Phase I	NCT02237638	Metastatic Cancer	30	Completed	Adverse events, Neutralization of endogenous VEGF in serum, Maximum tolerated dose	2014/4
IL-2 secreting autologous neuroblastoma vaccine	Phase I	NCT00048386	Neuroblastoma	13	Completed	Antitumor response	1999/11
IL-2 stimulated NK cells	Phase I	NCT01147380	Liver Cirrhosis, Hepatocellular Carcinoma, Evidence of Liver Transplantation	18	Completed	Side Effect	2010/6
IL-2 gene-transduced autologous neuroblastoma cells	Phase I	NCT00186862	Neuroblastoma	24	Completed	Safety	1998/8
IL-2-secreting CD40L-expressing autologous B-CLL cells	Phase I	NCT00058786	Chronic Lymphocytic B-Leukemia	9	Completed	Safety	2002/12
ImmunoPulse IL-12®	Phase II	NCT01440816	Merkel Cell Carcinoma	15	Completed	Percentage of Participants Who experienced at least 2-fold increase in expression of IL-12 Protein	2012/1
ImmunoPulse IL-12®	Phase II	NCT01579318	Cutaneous T Cell Lymphomas, Mycosis Fungoides	2	Terminated	ORR	2012/6
ImmunoPulse IL-12®	Phase II	NCT02345330	Head and Neck Squamous Cell Carcinoma	4	Terminated	Best Overall Response Rate (BORR)	2015/5
ImmunoPulse IL-12®	Phase II	NCT01502293	Melanoma	51	Completed	Best Overall ORR	2012/2
PENNVAX®-GP/ INO-6145 plus INO-9012	Phase I|Phase II	NCT03606213	HIV-1-infection	56	Completed	Adverse events, T cell responses	2018/8
GNOS-PV01 plus INO-9012	Phase I	NCT04015700	Glioblastoma	12	Recruiting	Safety, tolerability, feasibility	2020/7
K562/GM-CSF Cell Vaccine	Early Phase I	NCT00361296	Myelodysplastic Syndromes	9	Terminated	Hematologic Response Rate, Cytogenetic Response Rate	2007/9
Lethally Irradiated Lymphoma cells with GM-CSF K562 Cells	Phase I	NCT00487305	Follicular Lymphoma	40	Active, not recruiting	Safety, toxicity	2007/6
Allogeneic Neuroblastoma Tumor Cell Line Vaccines	Phase I|Phase II	NCT00703222	Neuroblastoma	7	Active, not recruiting	Safety	2008/6
Oncoquest-L	Phase II	NCT02194751	Follicular Lymphoma	30	Not yet recruiting	Overall tumor response rate	2021/7
Proscavax	Phase II	NCT03579654	Prostate Cancer	120	Not yet recruiting	Prostate Cancer progression	2019/2
MRC-5 transfected with DNA	Phase I	NCT00793208	Non Small Cell Lung Cancer	1	Terminated	Safety	2008/12
Vigil™	Phase II	NCT01453361	Advanced Melanoma	18	Terminated	Enzyme-Linked ImmunoSorbent Spot	2011/10
VXM01	Phase I	NCT01486329	Stage IV Pancreatic Cancer	72	Completed	Safety, tolerability	2011/12

4.1. Monotherapy using cancer vaccines

4.1.1. Targeting tumor vasculature

Angiogenic activity was higher in the poor prognostic group [156]. Due to the immunosuppressive function of tumor-associated angiogenesis and its crucial role in the development of tumors, antiangiogenic vaccines are a promising approach to reduce tumors without significant toxicity. Notably, antiangiogenic therapy has an innate advantage in that agents can directly contact targets in the circulation as soon as they are injected.

Antiangiogenic therapy creates a transient window of vessel normalization and has immune potency for reprogramming the TME away from immunosuppression [157]. Many factors are involved in the formation of tumor vasculature. Among them, VEGF, one of the proangiogenic factors, is the most studied growth factor in the anti-angiogenesis field. VEGF is strongly expressed on tumor endothelial cells compared to noncancerous tissue [158,159]. Currently, many studies targeting VEGF are underway. For instance, the listeriolysin O–Flk1 (LLO-Flk1) vaccine targets Flk-1, which is the mouse homolog of VEGF reporter 2, a strong therapeutic target. It has been found to be involved in mediating tumor vessel destruction without causing significant side effects during pregnancy and wound healing, suggesting important roles in targeting tumor vasculature-related factors [160,161].

Basic fibroblast growth factor (bFGF), also called FGF-2, is another angiogenic factor that modulates mitosis and angiogenesis. In melanoma, FGF-2 is uniquely expressed on malignant cells, endowing it with high efficacy and safety as a promising target against melanoma [162]. Although autologous FGF-2 may be identified as a self-molecule, the human recombinant FGF-2 protein vaccine as an exogenous antigen can break self-tolerance and tends to have more prominent potency [163]. Additionally, it has also been reported that recombinant FGF-2 recombinant is effective in inhibiting metastasis of both melanoma and lung cancer, further suggesting the vital role of targeting angiogenesis-related factors in metastasis [164]. The same effect was also presented in another study for a liposomal FGF-2 inhibitory peptide vaccine [165].

Another endothelial antigen, tumor endothelial marker 1 (TEM1/CD248), is expressed by tumor vessel-associated pericyte cells rather than normal vessels [166]. Due to its significant role in tumor angiogenesis and growth and its negative correlation with the clinical outcome of cancer patients, TEM1 may generate efficient therapeutic effects with minimal toxicity [167]. Likewise, a DNA vaccine called TEM1-TT also controlled the progression of established tumors in a mouse model. Notably, TEM1-TT did not mediate wound healing or pregnancy, although neoangiogenesis participates in these two processes, indicating its safety [168].

In addition to targeting growth factors involved in the regulation of angiogenesis, another direction is targeting the vasculature itself. SANRAVAC is a set of all natural target antigens for vaccination against cancer authenticated from the cell surface with an advanced proteomics approach. Its core mechanism of action depends on the heterogeneity of endothelial cells [169,170]. In vitro data show that autogenic SANTAVAC can destroy 18 tumor endothelial cells before 1 normal endothelial cell is destroyed by virtue of a highly specific antigen composition code [171]. It also destroyed 17% of the targeted tumor vasculature before normal cells were adversely affected in another study [172,173]. This kind of vaccine is also designed to eliminate tedious steps in which staff collect patient biomaterials and process them to available antigens. This technology facilitates translation into clinical practice.

4.1.2. Targeting CAFs

As the dominant cell type in the solid TME mentioned before, CAFs have been highlighted to play a promising role as therapeutic targets for the improvement of immunotherapeutic approaches against cancer.

Attempts to therapeutically target CAFs have primarily focused on CAF-associated proteins. Fibroblast activation protein (FAP) is a specific protein overexpressed in CAFs and is expressed at low levels or absent in normal tissues, indicating that it represents an ideal universal stromal antigen. Prior studies have indicated its potential to affect the tumor stroma. More specifically, DNA and DC vaccines expressing human FAPα lead to preliminary attenuated expression of collagen I and other stromal factors in the TME [174], [175], [176]. The FAP DNA vaccine also leads to a proimmune shift of Th cells in the immune microenvironment from Th2 to Th1 polarization in breast cancer [177,178]. A synthetic consensus (SynCon) sequence approach was designed to maintain FAP protein structure while introducing new neoepitopes. This SynCon FAP DNA vaccine supports breaking tolerance in lung and prostate tumor models [179]. However, a FAP vaccine using a recombinant adenovirus vector failed to induce effective antitumor benefits. High expression of the cytokine IL-10 is considered the primary cause, indicating the necessity of combining multiple cytokines or proteins associated with CAFs in cancer treatment [180].

TGF-β is another major CAF protein. The TGF-β antisense vaccine significantly enhances tumor cell immunogenicity in a glioma mouse model [181]. Subsequently, a clinical phase II study of NSCLC showed similar results [182]. Unfortunately, a subsequent phase III trial did not display a positive overall survival outcome, and the underlying reason is to be studied [183].

4.1.3. Vaccines with cytokine adjuvants

Admittedly, TME-targeting vaccines have unsatisfactory immunogenicity due to technology limitations. Currently, recombinant vaccines composed of TAA targeting malignant cells and adjuvants improving the immune TME are more common and potent. Of note, some of them are briefly described in the previous sections. This section will discuss several primary cytokines applied in vaccination, including IDO1, GM-CSF, IFN-γ, interleukins and BAFF.

4.1.3.1. IDO1

Mouse IDO1-derived peptides are capable of eliciting CD8⁺ and CD4⁺ T-cell-mediated antitumor responses based on the expression of IDO1, which localizes in infiltrating immune cells [184]. A clinical phase I vaccination trial (NCT01219348) using an IDO-derived A2 peptide has been performed on 15 advanced non-small-cell lung cancer (NSCLC) patients. Excitingly, the results showed a higher median overall survival (OS) (>2 years), as well as a reduced Treg population and no severe toxicity in corresponding patients [185,186].

4.1.3.2. GM-CSF

Among all cytokines, GM-CSF seems to be the best candidate for adjuvant treatment in immunotherapy. GVAX vaccines are comprised of genetically modified tumor cells engineered to secrete GM-CSF. Currently, GVAX vaccines have been studied in more than 600 patients. Early phase clinical trial data suggest that GVAX cancer vaccines are well tolerated, showing objective evidence of antitumor activity in multiple tumor types [187].

4.1.3.3. IFN-γ

IFN-DCs, which are mentioned before, exhibit more advanced maturation than other DCs. They have the features of both plasmacytoid DCs and NK cells [182]. The advantage of IFN-DCs is their higher efficiency in targeting antigens onto MHC I compared to their normal mature IL-4⁺ DC counterparts [188]. Meanwhile, IFN-γ from IFN-DCs stimulates systemic cytokine secretion, thereby enhancing efficacious and safe antitumor immunity. For example, IFN-DCs loaded with an immunogenic TCL elicit high titres of IFN-γ and in turn induce a systemic Th1-Th17-skewed immune response in lymphoma [189].

4.1.3.4. Interleukins

A renewed interest in interleukin-based drugs has led to an exponential increase in the number of clinical trials exploring their safety and efficacy, not only as traditional agents but also as cancer vaccines [190]. IL-15 improves vaccine immunity due to its ability to support the functionality of DCs, B cells, T cells and NK cells and to rescue tolerant or dysfunctional CD8⁺ T cells [191]. DC-derived IL-15-based DNA vaccines (IL-15⁺ DCs) displayed better feasibility by virtue of low toxicity than rIL-15, indicating a better therapeutic efficacy of vaccines with interleukins as adjuvants [192]. Analogously, vaccines modified with IL-24 or IL-10 also exhibited decent antitumor activity [193], [194], [195]. Recently, IL-18 has shown great anti-immune potential as an inflammasome-mediated cytokine, suggesting that vaccines with IL-18 may be an effective strategy for successful therapeutic vaccination. For example, administration of IL-18/gp100-DCs successfully induces infiltration of inflammatory cells inside and around the tumors, followed by the significantly increased activity of NK cells and the promoted production of Th1 cytokines (IL-2 and IFN-γ), indicating a strong activation of the antitumor immune response [196]. Moreover, the addition of a plasmid vector equipped with HSV-TK and IL-18 driven by the hTERT promoter also contributes to reliable immune potency [197].

4.1.3.5. BAFF

BAFF is an immunostimulant, as mentioned above [31,126]. However, systemic administration of BAFF is nearly impossible due to the wide expression of its receptors on a variety of cells. Activin A (ActA)-treated DCs represent a novel opportunity to stimulate BAFF. ActA triggered the SMAD2 and ERK1/2 pathways in DCs and upregulated the production of BAFF and APRIL. Prevention of BAFF and APRIL production in ActA-DCs completely abrogated the upregulation of the antitumor potential of DCs, suggesting that local delivery of these cytokines by DCs led to an augmented antitumor immune response [198].

Taken together, this section highlights the utility of vaccines to target the TME for further exploration and refinement. It has been discovered that the efficacy of some vaccines can be significantly improved by targeting components of the TME directly or by improving the TME through the addition of adjuvants. Therefore, refinement of the TME may be a promising strategy for vaccine design and therapy.

4.2. Combination therapy

Although therapeutic vaccines show decent potency due to the refinement of the TME in some cases, it is still common that most cancer vaccines currently show minimal clinical efficacy. One of the plausible explanations is that vaccine-activated T cells are maintained in a tumor-induced immunosuppressive environment. Hence, combination with other therapeutics seems to be a good choice (Table 4). Currently, some coping strategies have emerged, such as ICB therapy and chimeric antigen receptor (CAR)-T-cell-based treatment [199]. They have the potential to enable the cytotoxic effects of vaccine-activated T cells by inhibiting signaling pathways mediating immunosuppression.

Table 4.

Combination therapy with a cancer vaccine targeting the TME.

Cancer vaccine	Combination	Phase	NCT	Disease	N	Recruitment status	Primary outcome measures	Study start date
Allogeneic GM-CSF secreting cellular vaccine	Docetaxel	Phase III	NCT00133224	Prostate Cancer	408	Terminated	Survival	2005/8
Allogeneic GM-CSF-secreting breast cancer vaccine	Trastuzumab, Cyclophosphamide	Phase II	NCT00399529	Breast Cancer	22	Completed	Safety, Clinical Benefit	2006/9
Allogeneic GM-CSF-secreting breast cancer vaccine	Trastuzumab, Cyclophosphamide	Phase II	NCT00847171	Breast Cancer	20	Completed	Safety, CD4+/CD8+ T-cell responses	2008/12
Allogeneic GM-CSF-secreting breast cancer vaccine	Trastuzumab, Cyclophosphamide	Phase II	NCT00971737	Breast Cancer	63	Completed	Toxicity, Clinical Benefit	2009/7
Allogeneic GM-CSF-secreting breast cancer vaccine	Cyclophosphamide, Doxorubicin hydrochloride	Phase I	NCT00093834	Breast Cancer	60	Completed	Toxicity, Immune responsesImmu'ne	2004/1
Allogeneic GM-CSF-secreting breast cancer vaccine	Recombinant interferon alfa, Cyclophosphamide	Phase I|Phase II	NCT00095862	Breast Cancer	24	Terminated	Safety, Tolerability, Feasibility	2004/11
Allogeneic Myeloma Vaccine	Lenalidomide	Phase II	NCT01349569	Multiple Myeloma	19	Completed	Response Conversion Rate	2012/1
Autologous, pancreatic Cancer Cells	Cyclophosphamide	Phase II	NCT01088789	Pancreatic Cancer	72	Recruiting	DFS	2010/4
CD40L and IL-2-expressing B-CLL cells	Lenalidomide	Phase I	NCT01604031	B-cell Chronic Lymphocytic Leukemia	2	Terminated	Safety	2013/2
Colon GVAX	Cyclophosphamide	Phase I	NCT00656123	Metastatic Colorectal Cancer	9	Completed	Safety	2008/3
GM-CSF vaccine	Lenalidomide	Phase II	NCT03376477	Multiple Myeloma	54	Recruiting	PFS	2019/9
GVAX	Cyclophosphamide, Pembrolizumab	Phase II	NCT02981524	Metastatic Colorectal Cancer	17	Completed	ORR	2017/5
GVAX	Cyclophosphamide, SGI-110	PhaseI	NCT01966289	Metastatic Colorectal Cancer	18	Completed	Difference in CD45RO+ TILs, Safety, Tolerability	2014/4
GVAX pancreatic cancer vaccine	Adjuvant Chemoradiotherapy	Phase II	NCT00084383	Pancreatic Cancer	60	Completed	OS, DFS	2002/1
GVAX pancreatic cancer vaccine	Cyclophosphamide	Phase II	NCT00727441	Pancreatic Cancer	87	Completed	Safety	2008/7
GVAX vaccine	Nivolumab, Ipilimumab	Phase I	NCT04239040	Neuroblastoma	26	Recruiting	Safety, Feasibility	2020/1
HLA-A*02:01-restricted VEGFR1-derived peptide vaccination	Gemcitabine	Phase I|Phase II	NCT00683085	Advanced Pancreatic Cancer	2	Terminated	Safety	2008/5
Trivax	Temozolomide, Radiotherapy	Phase II	NCT01213407	Glioblastoma Multiforme	87	Completed	PFS	2010/4
INO-3112	Durvalumab	Phase II	NCT03439085	Recurrent or Metastatic Human Papillomavirus Associated Cancers	77	Active, not recruiting	Objective Response Rate (ORR)	2018/11
Pancreatic GVAX	Cyclophosphamide, Cetuximab	Phase II	NCT00305760	Pancreatic Cancer	60	Completed	Safety	2005/12
Melanoma GVAX	Cyclophosphamide	Phase I	NCT01435499	Melanoma	21	Completed	Safety, Tolerability	2011/9
MVA-MUC1-IL2	Cisplatin, Gemcitabine	Phase II|Phase III	NCT00415818	Advanced Non-Small Cell Lung Cancer	148	Completed	PFS	2005/12
Pancreatic GVAX	Ipilimumab	Phase I	NCT00836407	Pancreatic Cancer	30	Completed	Toxicity	2009/2
Pancreatic GVAX	Ipilimumab, FOLFIRINOX	Phase II	NCT01896869	Metastatic Pancreatic Adenocarcinoma	83	Completed	OS	2013/11
Pexa-Vec	Ipilimumab	Phase I	NCT02977156	Metastatic/Advanced Solid Tumor	22	Active, not recruiting	Dose Limiting Toxicities, ORR	2017/1
PSA-TRICOM	Enzalutamide	Phase II	NCT01875250	Prostate Cancer	38	Completed	Tumor Growth Rate	2013/7
Survivin Vaccine	Autologous hematopoietic cell transplantation, Prevnar 13, G-CSF	Early Phase I	NCT02851056	Multiple Myeloma	14	Active, not recruiting	CR and Safety	2016/9
Tecemotide	Bevacizumab, chemoradiotherapy, chemotherapy	Phase II	NCT00828009	Lung Cancer	70	Completed	Safety	2010/12
Unmodified SKNLP, with gene-modified SJNB-JF-IL2 and SJNB-JF-LTN neuroblastoma cells	Cytoxan	Phase I|Phase II	NCT01192555	Neuroblastoma	11	Active, not recruiting	Safety	2010/9
VEGFR1-A24-1084 (SYGVLLWEI)	Gemcitabine	Phase I|Phase II	NCT00683358	Advanced Pancreatic Cancer	14	Unknown	Phase I: Safety; Phase II: Time to Progression (TTP)	2008/5
Vigil™	Carboplatinum, Taxol	Phase II	NCT01867086	Ovarian Cancer	1	Completed	TTP, Response Rate	2013/6
Vigil™	Bevacizumab	Phase II	NCT01551745	Ovarian Cancer	5	Completed	TTP, Response Rate	2012/3

4.2.1. Combination with anti-PD-(L)1 therapy

It is well known that anti-PD-(L)1 therapy effectively improves the immune TME. Therefore, the combination of ICBs and vaccines is considered a feasible strategy to this end. Preclinical studies have shown that combined tumor antigen-loaded DC vaccines and anti-PD-1 therapy exerted a strong synergistic effect in many cancers, including lung cancer, breast cancer, melanoma and hepatocellular carcinoma [200], [201], [202], [203], [204].

Trp2 is a melanoma antigen peptide. As reported, a combination of ICB with a TRP2 mRNA vaccine led to the inhibition of tumor growth in a melanoma mouse model by downregulating PD-L1 in DCs [205]. Similarly, combined therapy harnessing anti-PD-L1 therapy and Trp2 peptide vaccines also improved the immune microenvironment in B16 melanomas, implying a potentially good prognosis [206]. In addition, PAS vaccination targeting gastrin, which is related to tumor-associated fibrosis and tumor immunity, induced a synergistic effect to reduce tumor growth and improve the T-cell response when combined with PD-1 Ab therapy in pancreatic cancer, while monotherapy failed to work [207].

Interestingly, the combination of ICBs with vaccination can also be presented by loading PD-1/PD-L1 on vaccines. Tumor cell vaccines secreting PD-1 neutralizing antibodies and GM-CSF induced remarkable antitumor immune effects and prolonged the survival of tumor-bearing animals [208]. Self-antigens may escape immunity due to central and peripheral tolerance mechanisms, including PD-L1. The nitrated T-cell epitope is a universal T helper epitope containing p-nitrophenylalanine, which is an immunogenic unnatural amino acid. Therefore, it can form a “nonself” epitope to break immune tolerance [209]. The PDL1-NitraTh vaccine can enhance the ADCC activity of antibodies induced by upregulation of PD-L1 [210].

Of note, there are some clinical trials trying to evaluate the potential of such combination strategies for cancer treatment. One of them is the NCT02426892 trial, which was a randomized, single-arm phase 2 clinical trial study to test the efficacy of ISA101 (HPV vaccine) combined with nivolumab (PD-1 inhibitor) for incurable HPV-16-positive solid tumors [211]. In this study, 34 patients were recruited, and ORR improvement was evident compared to PD-1 blockade alone (33% for combination vs 16%-22% for alone). These results showed that combination strategies could be a feasible treatment for cancers.

However, many problems remain unsolved. For instance, it seems that the combination strategy is more suitable for DC vaccines than peptide vaccines. One explanation is that peptide vaccines induce T cells with lower cytotoxic activity than DC vaccines do, limiting their combination with other immunotherapies. Correspondingly, a study showed that only DC vaccines rather than peptide vaccines combined with anti-PD-1 therapy augmented antitumor activity [212]. As a result, most PD-(L)1 blockade-based combination therapies focus on DC carriers.

4.2.2. Combination with CTLA-4

Similar to PD-1, activation of another checkpoint named antigen-4 (CTLA-4), which is expressed on the surface of T cells, contributes to the inhibition of T cells. While CTLA-4 blockade enhanced the priming of responsive T cells, it did not prevent the induction of tolerance to tumor antigens [213]. Combining vaccines with anti-CTLA-4 therapy will break this deadlock, improving the therapeutic effect by enhancing the levels of T-cell activation and memory [214].

Numerous clinical trials have been performed to explore the feasibility of treatment using vaccines in combination with CTLA-4 blockade.

In a phase II study (NCT01302496), 39 patients with pretreated advanced melanoma were enrolled. These patients were treated with a mRNA DC vaccine (TriMixDC-MEL) plus ipilimumab (anti-CTLA-4 monoclonal antibody) as the first-line therapy. Specifically, they received four cycles of combination treatment in total, followed by maintenance therapy [215]. Here, the 6-month disease control rate (DCR) was 51% (19/39). The overall tumor response rate was 38%; 8 patients achieved a complete response, and 7 patients achieved a partial response. The most common grade 3/4 AEs were hepatitis (5/39, 12.8%) and hypopituitarism (3/39, 7.7%). Notably, the DCR firmly exceeded monotherapy in other prospective clinical trials [216], [217], [218]. This finding supports the synergistic effects of DC vaccines and anti-CTLA-4 immunotherapy, which could be a promising choice for patients with pretreated advanced melanoma.

Compared to single-arm clinical trials, multiarm trials seem more convincing. A phase III study (NCT01302496) indicated that ipilimumab has improved efficacy with the gp100 peptide vaccine in melanoma. Gp100 is a common melanoma antigen. In this study, a total of 676 patients were enrolled (403 patients in the combination group, 137 patients in the ipilimumab monotherapy group and 136 patients in the vaccine monotherapy group). The median OS was significantly longer in the combination group than in the vaccine monotherapy group (10.0 months vs. 6.4 months). However, the median OS was almost nondistinctive between the combination and ipilimumab monotherapy groups (10.0 months vs. 10.1 months). Generally, this trial indicates that anti-CTLA-4 therapy is beneficial to vaccination [219].

4.2.3. Combination with CAR-T cells

CAR-T-cell immunotherapy has achieved a great success in hematological cancers, but it has difficulty in the treatment of solid tumors due to the immunosuppressive TME [220]. Therapeutic vaccination is one approach for enhancing endogenous T-cell responses against cancer, which is likely to partially lower the effects of the immunosuppressive TME on CAR-T-cell therapy. Hence, it is potentially feasible to combine vaccination and CAR-T immunotherapy. Amphiphiles (amph vaccines) aim to deliver vaccines to lymph nodes by linking peptide antigens to albumin-binding phospholipid polymers. This strategy led to a 30-fold increase in T-cell priming and then enhanced antitumor efficacy with a great reduction in systemic toxicity [221]. Furthermore, this team developed an amph vaccine linked to EGFRvIII (amph-pepvIII) to improve CAR-T-cell proliferation, as well as their infiltration into tumor sites. To achieve a better outcome, another research and development team further designed a bispecific CAR platform to overcome the limitation of EGFRvIII as a line epitope. FITC/TRP1–CAR-T cells can be activated by either amph-FITC-coated target cells or TRP1-expressing B16F10 cells. Therefore, amph-FITC vaccination substantially stimulated the proliferation of FITC/TRP1-bispecific CAR-T cells. Their combination obviated the need to develop a vaccine individualized for each target antigen, greatly reducing the cost [222].

Certainly, researchers are also seeking other promising targets that can help enhance the potency of CAR-T therapy. CLDN6, an oncofetal cell-surface antigen, is a potential candidate that appears suitable for CAR-T-cell targeting. It has been reported that CAR-T cells activated by CLDN6 are strong IFN-γ-secreting effectors [223]. Accordingly, a CAR-T-cell-amplifying RNA vaccine (CARVac) encoding a chimeric receptor directed toward CLDN6 was designed for body-wide delivery of the CAR antigen into lymphoid compartments, potentially improving the antitumor efficacy of CAR-T cells. This lipoplex RNA vaccine contained natively folded CLDN6 on the surface of DCs, which in turn enhanced the efficacy of adoptively transferred CLDN6-CAR-T cells and enabled therapeutic tumor control at a lower dose of CAR-T cells [224]. The safety and efficacy of this combination are currently being evaluated in a phase I/II clinical study (NCT04503278).

4.2.4. Combination with targeted therapy

In regard to targeted therapy, HER2-based treatment stands out. Trastuzumab (anti-HER2 mAb) in combination with a GM-CSF-secreting breast tumor vaccine showed safety and a trend toward longer PFS and OS at 7 and 42 months, respectively [225]. Interestingly, most drugs that target HER2 are known to downregulate VEGF expression as well [226,227]. These observations led to the hypothesis that targeting both HER2 and VEGF might exert synergistic antitumor effects. It has been confirmed that the combination of the HER2 peptide epitope and VEGF peptide mimic (RI-VEGF-P4CYC) induces potent antitumor and antiangiogenic responses in vivo [228].

The combination of anti-CD4 mAbs may provide a new choice for cancer vaccine-based therapy. CD4 depletion performed by anti-CD4 mAb significantly reduced Treg cells when combined with multiple vaccinations, finally leading to the recovery of therapeutic efficacy [141]. Another study showed that the use of anti-CD4 mAb in combination with OVA peptide vaccine therapy resulted in an increased therapeutic effect [229]. Similarly, the anti-CD40 antibody mitazalimab was found to be involved in the activation of APCs and depletion of Tregs, enhancing antitumor efficacy and prolonging the survival of mice bearing MB49 bladder carcinoma tumors when combined with the OVA peptide vaccine [230].

Targeted therapy also contributes to the lifespan extension of DCs, which reverses the immunosuppressive TME in response to vaccination. In fact, induced activation of DCs shortens the DC lifespan, which is associated with metabolic transition due to the regulation of mammalian target of rapamycin (mTOR) [231]. Therefore, it is reasonable to increase antitumor activity by targeting mTOR. It has been reported that tumor-bearing mice treated with the CTGF/E7 DNA vaccine combined with a mTOR inhibitor exhibited higher percentages of mature DCs in the TME, better disease control and longer survival [232].

In total, targeted therapy can either help eliminate tumor cells directly or activate immune cells, thus increasing the likelihood of success of the concomitant cancer vaccine.

4.2.5. Combination with chemotherapy

Chemotherapy has long been considered immunosuppressive. However, increasing data indicate that chemotherapy promotes tumor immunity through ICD [233,234]. Moreover, chemotherapy can convert the suppressive TME by elevating antigen exposure [235]. Several preclinical and clinical studies have shown the promising prospects of combining cancer vaccines and chemotherapy. For example, lenalidomide plus an IFN-DC vaccine or DNA vaccine exerted an additive therapeutic effect in lymphoma [236,237]. Similarly, gemcitabine treatment with HER2 DNA vaccines enhanced the sensitization of tumor cells to CTL-mediated killing [238]. Likewise, pomalidomide, as an analog of lenalidomide, combined with TAA DCs and dexamethasone synergistically enhanced antimyeloma immunity by skewing the immunosuppressive status toward an immunosupportive status [239].

Cyclophosphamide (CY) remains one of the most successful and widely utilized antineoplastic drugs. As mentioned above, IL-10 played a part in limiting CAF vaccine application [180]. CY downregulated IL-10, showing the potential for combination with FAPα-targeted vaccines [240]. The combination of CY with a FAPα-targeted vaccine reduced both the quantity of Tregs and the expression of IL-10, in turn leading to tumor inhibition by 90.2% and survival extension by 35% in a murine model [241]. Apart from CY, doxorubicin combined with FAPα/survivin double-target DNA vaccines can also inhibit lung metastasis and prolong the survival of mice by reducing CAF-related cytokines and chemokines [242]. Overall, the GM-CSF vaccine plus CY agents can be tolerable and effective in curing HER2-positive metastatic breast cancer.

Furthermore, similar results were achieved in a study of a combination strategy for HPV vaccines. The therapeutic effect of an HPV vaccine was little pronounced in a genital tract model due to high infiltration with Tregs and MDSCs. Notably, such a hostile tumor microenvironment can be reversed by cisplatin treatment, leading to a complete regression of tumors when combined with mRNA vaccines [243]. Similarly, an open-label observational study also indicated the potential of such a combination strategy, which enrolled 19 advanced cervical cancer patients (6 patients in the combination cohort and 13 patients in the vaccine cohort). This trial revealed that carboplatin-paclitaxel therapy fostered HPV vaccine-induced T-cell responses and reset myeloid cell depletion [244]. However, negative results should not be overlooked. A prospective, randomized phase 2 trial (NCT02285413) tested the immunogenicity and feasibility of an autologous DC vaccine combined with cisplatin in patients with stage III and IV melanoma (22 stage III patients and 32 stage IV patients). Four patients stopped cisplatin and then continued DC monotherapy due to toxicity. In terms of tolerability, 1 of 54 patients experienced grade 3 decompensated heart failure. Of note, this combination did not result in more tumor-specific T-cell responses or improved clinical outcomes compared to monotherapy [245]. Therefore, the feasibility of such combination strategies needs to be further evaluated.

4.2.6. Combination with radiotherapy

Radiotherapy (RT) can stimulate an immunosupportive response by ICD-induced abscopal effects [246]. Interestingly, local low-dose tumor irradiation influences the intratumor immune cells and chemokines induced by vaccines rather than whole-body irradiation [247,248].

As is widely known, RT induces tumor vasculature damage, especially with a single large dose of radiation (<10 Gy) [246,249]. TEM1 is an endothelial antigen as mentioned above. It has been reported that the addition of RT to heterologous TEM1 vaccination (anti-vasculature antigen vaccine) reduced the progression of irradiated and abscopal distant tumors compared to vaccine treatment alone. The major explanation was probably that RT-induced MHC-I expression on endothelial cells improved the immune recognition of the endothelium by anti-TEM1 T cells with subsequent severe vascular damage, potentially enhancing the antitumor response [250].

In an observational cohort study, DC vaccines plus RT served as a treatment for 40 patients with esophageal cancer [251]. Assigned with heat shock protein (HSP)-loaded DC vaccination (4 injections) and RT (60 Gy/in 5 fractions per week for 6 weeks), the levels of Th1 cytokines (IL-2, IL-12, and IFN-γ) and antigen-specific IFN-γ⁺ CD8⁺ T cells were increased compared to baseline and the control group (+92.4%, +70.9%, 214.3% and 3-16.4-fold, respectively). No severe AEs or AE-related discontinued treatment was observed. According to the cutoff data, this combination regimen exhibited promising tolerability. Nevertheless, further evaluations are necessary for confirmation of its efficacy and safety in the long run.

In addition, the multicohort phase Ib trial (NCT01915524) tested the therapeutic effect of the BI1361849 (CV9202) mRNA vaccine together with RT (4 × 5 Gy) in 32 stage IV NSCLC patients (six patients failed screening) [252]. In this study, the stable disease rate was 46.2%, and the increased rate of BI1361849 antigen-specific immune responses was 84%. Regarding toxicity, four patients (15.4%) experienced grade 3 AEs. This evidence supports further investigation of CV9202 in combination with RT.

In summary, a cancer vaccine targeting the TME is an effective approach to mobilize exterior components to kill malignant cells. Combinations with other therapies can stimulate immune cells to achieve maximal efficiency.

5. Conclusion and perspectives

Therapeutic cancer vaccines have raised many hopes for maximizing active immunity. Early development and application of cancer vaccines focuses on TAAs or TSAs. As self-antigens, TAA/TSA vaccines easily escape the immune response, and a series of associated clinical trials ended with failure [4,10,12]. In fact, the key indicator for a successful cancer vaccine is boosting the immune system for antitumor response. Extensive evidence has pointed to the complicated relationship between the TME and vaccine efficacy. On the one hand, the TME is critical for vaccine efficacy, which partially explains why different individuals with a suppressive TME are more likely to be insensitive to vaccine therapy. On the other hand, vaccines can reshape the TME, which in turn influences their potency. After vaccination, not only the immune landscape but also the metabolic and vasculature environments are changed, resulting in the conversion of the original TME into a ‘hot’ or a ‘cold’ TME. Overall, the TME impacts the therapeutic potency of cancer vaccines, and vaccines targeting the TME provide new hope for revival in the cancer vaccine field.

Despite promising prospects for vaccines, the current situation is not optimistic and is worth pondering. Many incompletely understood mechanisms can lead to therapeutic resistance to cancer vaccines with low immunogenicity. This resistance mainly includes two types, namely, innate resistance and acquired resistance. Innate resistance might be caused by some immunosuppressive components, such as suppressive immune cells and cytokines, in certain subpopulations of patients. For acquired resistance, induced immunosuppressive molecules and signaling pathways, such as immune checkpoints and IFN-γ, might be important. As a result, the suppressive TME is a major challenge to break the deadlock for both types. To thwart therapeutic resistance, different supporting therapies, such as chemotherapy, targeted therapy, radiotherapy and other immunotherapies, will need to be administered in combination with cancer therapeutic vaccines. Novel delivery platforms, as well as the addition of several adjuvants, have also advanced to optimize vaccine potential [253,254]. LNP is the most studied carrier due to its low toxicity and robust carrying capacity [255].

Notably, the timing, sequence and manner of different treatments need to be carefully determined. The components of the TME can change over time, exerting an antitumor effect at an early oncogenesis stage but acquiring a protumoral effect at an advanced stage. For instance, patients with advanced melanoma exhibit reduced immunologic efficacy in response to vaccines due to increased Tregs [256]. Current vaccine injections vary, mainly including intradermal (i.d.), subcutaneous (s.c.), intravenous (i.v.), intraperitoneal (i.p.), intralymphatic and intratumoral injection. Current evidence supports that DC given s.c. is superior to DC given i.v. or intratumorally [257]. Therefore, choosing the appropriate time and injection route may increase the chance of clinical benefit [258].

The challenges surrounding the use of cancer vaccines are nonnegligible. TME vaccines have achieved rapid development, especially in anti-CAF vaccines and antivessel vaccines, but most of them function as adjuvant vaccines. Cancer vaccines also need biomarkers similar to other immunotherapies, although related research is rare. It has been reported that immunosuppressive myeloid cells are negatively correlated with HPV vaccine potency in cervical cancer patients, suggesting their potential role as a predictive biomarker for vaccine efficacy [259]. Preclinical research reported an inverse correlation between cancer vaccine efficacy and TILs. Thus, the Vx-001 vaccine is clinically active only in nonimmunogenic/cold NSCLC [260]. Consequently, TME regulation still works as an auxiliary component in vaccines. With further understanding of bidirectional communication between the tumor and the TME, future studies will explore cancer vaccines by directly targeting immune pathways in the TME to potentiate vaccine outcomes.

Abbreviations

ActA

Activin A

BAFF

B-cell–activating factor

bFGF

Basic fibroblast growth factor

CIITA

Class II transactivator

COX

Cyclooxygenase

CRCL

Chaperone-rich tumor-cell lysate

DCR

Disease control rate

ECM

Extracellular matrix

GSK-3β

Glycogen synthase kinase-3β

HSP

Heat shock protein

ICD

Immunogenic cell death

ISVs

In situ cancer vaccines

LLO-Flk1

The listeriolysin O–Flk1

LNP

Lipid nanoparticle

LPS

Lipopolysaccharide

PAMPs/DAMPs

Pathogen/damage-associated molecular patterns

PDT

Photodynamic therapy

PRRs

Pattern recognition receptors

PS

Phosphatidylserine

SANTAVAC

The set of all natural target antigens for vaccination against cancer

SK

Shikonin

SynCon

Synthetic consensus

TCL

Tumor cell lysate

TEM1

Tumor endothelial marker 1

TLRs

Toll-like receptors

Declaration of competing interest

The authors declare that they have no conflicts of interest in this work.

Biographies

Siwei Zhang obtained her bachelor's degree in Central South University. She is now a doctor candidate in oncology at Shanghai Medical College of Fudan University. Her research focuses on the precision treatment of breast cancer.

Yi-Zhou Jiang (BRID: 03809.00.92827) is an associate professor of Department of Breast Surgery at Fudan University Shanghai Cancer Center. He also serves as the director of Discipline Construction Executive Committee at Fudan University Shanghai Cancer Center. He received his M.D. degree at Shanghai Medical College Fudan University in 2014. Focusing on molecular subtyping of breast cancer, he is carrying out a series of studies to advance the precision treatment, forming a closed-loop research system between fundamental research and clinical research. As the principal investigator, he has hosted over 10 research projects including National Natural Science Foundation. He has published papers in journals like Cancer Cell, Cell Metabolism, Cell Research, Molecular Cancer, Nature Communications, and Clinical Cancer Research.