Personalized Dendritic-cell-based Vaccines Targeting Cancer Neoantigens

Cancer Neoantigens as Targets

Tumor cells express mutant proteins that are normally not present owing to genetic and epigenetic alterations, presenting an altered repertoire of MHC-class-I- and MHC-class-II-related peptides. The activity of cytotoxic T cells against these MHC-class-I- and MHC-class-II-presented mutant antigens (i.e., neoantigens) may lead to cancer regression and is the subject of various basic and clinical studies. Neoantigens are produced by various genetic abnormalities, including somatic single-nucleotide variants (SNVs), gene fusions, and mutational frameshift insertions and deletions (indels) (9). In a mouse melanoma model, mutant synthetic long peptides (SLPs) generated by the transcription and translation of a mutant gene elicited strong cellular immune responses both in vitro and in vivo. Some of immunogenic SLPs exhibited a strong mutation-specific immune response with no cross-reactivity to wild-type peptides (10). Because neoantigens are derived from genetic mutations in tumor cells, they are present in tumor cells but not in normal tissues, unlike TAAs, which are present in both normal and tumor tissues (11). Therefore, neoantigens are ideal targets for cancer therapy that elicit a strong tumor-specific T cell immune response.

With the recent advent of next-generation sequencers (NGS), genetic mutations in individual patients can be easily detected, and artificial intelligence can predict antigens derived from genetic mutations, which have a high binding capacity to MHC molecules (10). Identifying neoantigens requires samples from both normal cells and tumor cells from the same patient. Whole exome DNA sequencing, which is used to detect nonsynonymous somatic mutations, can identify important variants in protein-coding regions. In addition, RNA sequencing can be used to identify expressed mutants and more accurately predict potential neoantigens (12). The bioinformatics pipeline for neoantigen prediction shares four major computational modules: HLA (human leukocyte antigen) typing, calling mutant peptides from somatic mutations and splicing variants, HLA binding prediction, and T cell recognition prediction. These technologies enable the design of neoantigen vaccines to elicit tumor-specific immune responses (13). Neoantigens are finally identified through several processes that validate the T cell response to the predicted antigen (Figure 1). The important function of neoantigens depends on three points: (i) tumor mutational burden (TMB) and neoantigens, (ii) presentation of neoantigens by MHC molecules, and (iii) recognition of cancer neoantigens by T cells. A positive correlation was observed between TMB and the number of cancer neoantigens (14). However, not all neoantigens can be further recognized by T cells and induce a subsequent immune response. Conversely, owing to tumor cell heterogeneity, even cases with a small TMB could exhibit stronger than average T cell reactivity (15).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Workflow for producing personalized DC-based neoantigen vaccines. Using tumor and normal tissues from patients, we can identify neoantigens produced by tumor mutations by analyzing differences in gene sequences between tumor and normal cells. HLA typing, identification of mutant peptides using a series of somatic mutations from whole exome sequencing of tumor and normal cell, HLA binding prediction, and tumor RNA sequencing data are conducted; in these analyses, various software packages to predict and prioritize immunogenic neoantigens are used. DC-based neoantigen vaccines are produced by loading DCs with SLPs or mRNAs based on the predicted neoantigens. DC-based vaccines are also produced by introducing tumor lysate or tumor mRNA extracted from tumor tissue into DCs by electroporation, and these personalized DC-based neoantigen vaccines are administered to patients to elicit a neoantigen-specific immune response. In addition, isolated tumor cells can be fused with DCs and administered to patients as DC/TC fusion hybrid vaccines. NGS: Next-generation sequencing; SLPs: synthetic long peptides; DC: dendritic cell; TC: tumor cell.

The efficient presentation of mutant peptides by MHC molecules is critical to their recognition by T cells and their subsequent activation, and although SNVs account for 95% of tumor mutations, indel or frameshift mutations generate immunogenic peptides through the juxtaposition of two peptide sequences or new peptide sequences. Owing to changes in an amino acid sequence and a spatial structure caused by indel or frameshift mutations, these mutant peptides have a stronger affinity for MHC molecules than mutant peptides of non-immunogenic SNVs that either bind poorly to MHC class I molecules or do not contain the mutated amino acid in the appropriate position and have been shown to be more likely to be recognized as neoantigens by T cells (16). T cells recognize antigens presented on MHC molecules, and CD8+ T cells are considered key effector cells in anti-tumor immune responses because of their important role in specifically recognizing cancer neoantigens (17). However, CD4+ T cells also play an important role in anti-tumor immunity as they assist in the initial stimulation and maintenance of CD8+ T cell responses (18), and CD4+ T cell responses have been shown to be essential for neoantigen-related tumor regression (19).

Neoantigens are generally classified into two subgroups: shared neoantigens and personalized neoantigens (8). Shared neoantigens may be common in some tumor types, whereas personalized neoantigens are antigens specific to each individual patient’s tumor. Therapy with personalized neoantigens offers a customized approach for each patient and may represent an important advance in cancer therapy, making neoantigen-targeted personalized immunotherapy the mainstream therapy. Neoantigen vaccines have been designed in various forms, including peptide, DNA, RNA, and DC-based vaccines (20). Among the various neoantigen vaccines being developed, peptide and mRNA neoantigen vaccines account for more than half of vaccines used in clinical trials (21).

In clinical trials of peptide vaccines, short-peptide vaccines were used in the early days, but SLP vaccines, which induce both CD4+ T cells and CD8+ T cells, are now the most common. The SLP vaccine platform has important advantages regarding the safety profile and the good manufacturing practice system. In addition, the 20-30 amino acid SLPs in cancer neoantigen vaccines may be rapidly processed and presented by APCs (22) and have the advantage of binding to both MHC class I and II molecules that activate CD8+ and CD4+ T cells, respectively. After cleavage and antigen processing of SLPs by proteasome, CD8+ T cells recognize short peptides of 8-11 amino acids bound to MHC class I molecules. In contrast, longer peptides of 14-18 amino acids bind to MHC class II molecules and are presented to CD4+ T cells (23).

mRNA neoantigen vaccines have unique advantages; for example, they are not incorporated into the host cell genome, and RNA can be extracted from small amounts of tumor tissue to be amplified for vaccine preparation when suitable tumor tissue is not available. Moreover, mRNA modifications and sequence optimization considerably increase the efficiency of protein expression from mRNA templates, making them more stable, durable, and easier to translate into encoded proteins (24). In terms of deliverability, upon intranodal injection of RNA, DCs in lymph nodes selectively uptake these RNA molecules encoding antigens, attract T cells, and induce expansion of antigen-specific CD8+ and CD4+ T cells (25). The ability to rapidly produce vaccines against arbitrary targets in dedicated mRNA production facilities is a major advantage of using mRNA in neoantigen vaccine development.

Principles of DC-based Vaccines

DC-based vaccines play an important role in cancer therapy: they induce the over-expression of MHC I and II and co-stimulatory molecules (such as CD80 and CD86), resulting in the effective antigen presentation to naive T cells (26). Co-culturing peripheral blood mononuclear cells (PBMCs) with granulocyte/macrophage colony-stimulating factor (GM-CSF) and interleukin 4 (IL-4) enhances the efficacy of antigen presentation of DCs (27). This protocol allows ex vivo cultured DCs to be used in clinical studies of cancer vaccines combined with various cancer antigens.

Various types of tumor-specific antigens, such as viral antigens and TAAs as well as neoantigens, including peptides, proteins and mRNAs, have been investigated using the DC-based vaccine platform (Figure 1). For DC/TC fusion vaccine, DCs can phagocytose isolated apoptotic TCs and acquire neoantigens upon fusion. In particular, anti-tumor peptides consisting of 9-15 amino acids are presented on MHC class I molecules on the surfaces of DCs and can be effectively stimulated by tumor-specific CD8+ cytotoxic T lymphocytes (CTLs). When extracellular protein antigens are used, antigens are presented via MHC class II and then via MHC class I by cross-presentation. Antigen presentation via MHC class I for DCs depends on antigen formulation for cross-presentation efficiency and the mode of antigen delivery during antigen processing. Antigen structure and loading method to DCs affect the efficiency of antigen presentation (28). However, loading antigens into DCs by electroporation could efficiently induce cross-presentation regardless of the antigen structures. Electroporation has been used for mRNA and protein loading into DCs, which has been shown to effectively induce antigen-specific CD8+ CTLs (29). Once immature DCs take up the antigen, they migrate to secondary lymphoid organs where they present the antigen to helper or effector T cells, triggering specific CTL responses (30).

DC-based vaccines have strong personalized vaccine features and are being investigated for targeting TAAs and/or neoantigens as individualized therapies. This is expected to elicit a strong immune response against patient-specific tumor antigens. DC-based neoantigen vaccines have shown potential as an effective and safe immunotherapy against solid tumors. Antigens containing neoantigens can be loaded into DCs by various methods, including the use of synthetic peptides, incubation with autologous tumor-derived whole mRNA and whole tumor lysates, and fusion with tumor cells. A key feature of many of these methods is the ability to pulse DCs with tumor lysates containing tumor-associated antigens to generate DC-based vaccines that harbor a wide range of tumor antigens, including neoantigens (31). Sipuleucel-T (Provenge^®, Dendreon Corp., Seattle, WA, USA), an FDA-approved autologous cellular immunotherapy for the treatment of metastatic hormone-refractory prostate cancer, uses APCs cultured with a recombinant fusion protein containing prostate acid phosphatase (PAP) to stimulate effector T cells (32). Previously, we successfully prepared an autologous tumor lysate-loaded DC vaccine with high cell viability via electroporation. Our study of the autologous tumor lysate-loaded DC vaccine confirmed its safety and feasibility for 41 patients with various solid tumors, such as carcinomas of the lung, colorectum, breast, stomach and liver. In addition, we observed no autoimmune reactions. The anti-tumor effect of the DC-based vaccine resulted in tumor regression or long-term stable disease, with tumor-reactive immunological responses for 13 patients (31.7%) (33). In addition, an autologous glioma lysate DC-based vaccine may be eligible for more glioma patients than a glioma-associated peptide DC-based vaccine since lysate is HLA-independent while peptide is HLA-dependent (34).

Although it is possible to generate a broad range of tumor-based antigens by utilizing fusion hybrid vaccines of DCs and tumor cells (DC/TC) as well as using tumor lysates (35), their indications are limited by the difficulty in obtaining tumor cells. However, in hematologic malignancies where tumor cells are readily available, such as multiple myeloma, DC/TC fusion hybrid vaccines may be able to induce tumor-specific immune responses and tumor regression (36).

mRNA transfection is a simple method to produce antigens in DCs. A Wilms’ tumor 1 protein (WT1) is reported to be over-expressed in the majority of acute myeloid leukemia (AML) cases. In the phase I/II trial of DC vaccines, DCs transfected with full-length WT1 mRNA for AML by electroporation yielded clinical responses that were found to be correlated with a vaccine-associated increase in WT1-specific CD8+ T cell frequency (37). Thus, a vaccine that introduces TAA mRNA by electroporation has already been generated, and specific immune induction and anti-tumor effects of mRNA-based DC vaccines have been confirmed in clinical trials. However, the development of DC vaccines using mRNAs encoding neoantigens predicted by NGS has not yet been reported. Currently, the most versatile neoantigen-specific DC vaccine platforms are those loaded with SLPs as the identified neoantigens (38).

Various attempts have been made to determine which form of neoantigen, such as SLPs, mRNA, protein, or whole tumor lysate is the best to load in DC vaccines, but no conclusions have yet been made. Furthermore, there are still many challenges, including DC maturation, as well as routes of administration to make DCs effective in priming and activating T cells in draining lymph nodes and the TME (3). Further research and clinical trials are needed to address these challenges, but DC-based vaccines hold much promise for cancer immunotherapy.

Personalized DC-based Vaccines Designed Without Neoantigen Identification by NGS

For DC-based vaccines, various tumor antigens are loaded into DCs; many previous clinical trials of DC vaccines have tested various TAAs along with neoantigens as personalized vaccines loaded into DCs. In an early clinical trial of a DC-based vaccine for glioma, patients received three treatments with a DC vaccine pulsed with tumor peptides eluted from primary cultured cells of an MHC-class-I-matched allogeneic glioblastoma. In this trial, the DC vaccine was well tolerated, and the patients showed a positive immune response, but no objective clinical response was observed (39). Although Sipuleucel-T for the treatment of prostate cancer was approved by the FDA in 2010 as an important milestone in the clinical application of DC vaccines, the efficacy of Sipuleucel-T has been limited, as its tumor progression-free effect has been inconsistent (29). In a study comparing the efficacy of an autologous DC vaccine with that of an autologous irradiated tumor cell vaccine (TCV), after five years of follow-up, these vaccine therapies were found to be well tolerated, and the DC vaccine was associated with a longer overall survival and a 70% reduction in risk of death (median survival was 43.4 months vs. 20.5 months for TCV) (40). In the case of treatment with ICT-107, which is an autologous DC vaccine pulsed with six different peptides targeting glioblastoma and cancer stem cells, the progression-free survival (PFS) in the ICT-107 group was significantly longer than that in the control group in HLA-A2+ glioblastoma patients with methylation of the O6-methylguanine-DNA methyltransferase promoter (24.1 vs. 8.5 months) (41). A significant association between immune responses and favorable clinical outcomes were also observed in these ICT-107 group patients.

The DC/TC fusion hybrid vaccine was well tolerated and safe for solid tumors and hematologic malignancies, such as AML and multiple myeloma. In addition, the DC/TC fusion vaccine increased both CD4+ and CD8+ T cells which express intracellular IFN-γ in response to in vitro exposure to tumor lysate. This suggests the broad antigen-presenting ability of the DC/TC fusion vaccine to produce TAAs containing various neoantigens recognized by the immune system (42).

In a study of DC vaccines in which mRNA was introduced by electroporation, DC vaccines loaded with whole tumor mRNA by electroporation improved the survival of immune responders with advanced melanoma, as confirmed by an immunological evaluation of the induction of T cell responses (43). In addition, DC vaccines against malignant melanoma transfected with mRNA encoding CD40L, caTLR4, and CD70 (TriMix mRNA) and mRNA for melanoma-associated antigen have demonstrated TAA-specific CD8+ T-cell skin infiltration (44).

The most important question is whether personalized DC vaccines loaded with tumor lysates or mRNA can induce T cell responses to tumor neoantigens arising from nonsynonymous somatic cell tumor mutations. Tanyi et al. reported that in a clinical trial of a DC vaccine pulsed with whole tumor lysates of ovarian cancer, neoantigens were identified by NGS in tumor samples from six patients, and T cell reactivity to neoantigens was examined for each patient after five DC vaccinations (45). Before DC vaccination, a neoantigen response was observed in only four patients, but after vaccination, it was observed in all six patients and the number of neoantigens to which T cells reacted increased. Neoantigen-specific CD8+ T cells isolated after vaccination were found to be multifunctional, simultaneously producing interferon (IFN)-γ, tumor necrosis factor (TNF)-α, and IL-2 in response to neoantigens. The results of immunological evaluations in this clinical trial demonstrated that the DC vaccine loaded with tumor lysates enhanced the specific response of T cells to neoantigens after personalized DC vaccination without the pre-identification of neoantigen candidates. Personalized DC vaccines loaded with tumor lysate or tumor mRNA, both containing neoantigens and TAAs, have been noted to potentially contribute to improving overall survival in malignant melanoma and glioblastoma (43, 46).

Clinical Applications of Personalized Vaccines Targeting Neoantigens Identified by NGS

Multiple clinical trials have shown that neoantigen-targeted vaccines are safe and induce specific T cell immune responses. The results of various basic studies and clinical trials indicate that future strategies for personalized neoantigen DC vaccines need to be developed. Significant benefits in terms of overall survival are still limited but are improving with the development of combined immunotherapies.

Neoantigen vaccines for SLP or mRNA platforms. Animal model studies have shown that neoantigens determined on the basis of MHC class I binding affinity are recognized by CD4+ T cells. Vaccines with such CD4+ immunogenic mutations have been shown to counteract the immunosuppressive TME and confer potent antitumor activity, providing the rationale for using SLPs that stimulate CD4+ and CD8+ T cells in neoantigen peptide vaccines (18). Individual HLA-binding neoantigen peptides selected on the basis of the predicted binding affinity were identified after surgery in six patients with resectable Stage IIIB or higher untreated high-risk melanoma. A total of seven subcutaneous inoculations of 13-20 SLPs based on the identified neoantigens were administered along with an adjuvant poly-ICLC. An enzyme-linked immunosorbent spot (ELISpot) assay detected multifunctional CD4+ T cells and CD8+ T cells, respectively, targeting 58 (60%) and 15 (16%) of all 97 unique neoantigens across all six patients. Immunological evaluation confirmed that the CD4+ T-cell response rate was higher than the CD8+ T cell response rate, even though neoantigens were prioritized on the basis of their predicted binding affinity to HLA class I molecules in humans. Four patients without metastatic disease had no recurrence after SLP vaccination. Two other patients with metastatic disease had disease progression after vaccination but received anti-PD1 checkpoint inhibitors and achieved tumor regression (47). Immunological responses of CD4+ T cells have been reported for neoantigen SLP vaccines in glioblastoma clinical trials. Thirteen newly diagnosed glioblastoma patients who were HLA-A*02:01- or A*24:02-positive were vaccinated with a conventional TAA (APVAC1), followed by a cancer neoantigen (APVAC2) SLP vaccine with the adjuvants poly-ICLC and GM-CSF. All the patients underwent surgical resection and received standard adjuvant chemotherapy with temozolomide postoperatively. APVAC1 induced a sustained CD8+ T cell response with a central memory phenotype. In contrast, APVAC2 elicited neoantigen-specific CD4+ T cell responses in all the patients, among which 84.7% (11/13) showed multifunctional CD4 responses with the Th1 phenotype (48). Another clinical trial of the neoantigen SLP vaccine for glioblastoma showed that the vaccine significantly increased the number of tumor-infiltrating lymphocytes (TILs) and induced potent multifunctional novel CD4+ and CD8+ T cell responses to neoantigens in patients who did not receive dexamethasone, which is given for brain edema induced in patients with glioblastoma (49).

In an mRNA neoantigen vaccine study, 13 patients with Stage III or IV malignant melanoma were treated with an RNA-based neoantigen vaccine prepared on the basis of an identified neoantigen. The RNA neoantigen vaccine was well tolerated with no serious adverse events. ELISpot assays of CD4+ and CD8+ T cells before and after vaccination showed that the immunoreactivity of T cells to neoantigens was enhanced in one-third of the patients and a novel immune response in the remaining two-thirds. CD4+ T cells, CD8+ T cells, and combinations of both cells were reactive to 57, 17, and 26% of neoantigens, respectively. Detailed analysis of postvaccination resected metastases from two of the three patients confirmed vaccine-induced T-cell tumor infiltration and neoantigen-specific tumor cell death. In addition, a comparison of the recurrence frequency in all patients recorded before and after neoantigen RNA vaccination showed a highly significant reduction in cumulative recurrent metastatic events over time (5).

DC-based neoantigen vaccines. In 2015, the first Phase I clinical trial of the first individualized neoantigen-loaded DC vaccine was conducted (50). The appropriate neoantigen was determined by epitope prediction using whole tumor exome sequencing in three patients with Stage III resected cutaneous melanoma who were previously treated with ipilimumab. Seven neoantigens selected from each patient and two gp100 peptides were loaded, and ex vivo cultured DCs were administered intravenously to the patients three times. After treatment, an enhanced immune response induced by T cells was observed, and all three patients survived. Moreover, no serious side effects were observed. This clinical study showed that the DC vaccine pulsed with individualized neoantigens was safe and well tolerated. In a study of personalized neoantigen-pulsed DC vaccines, 12 patients with advanced lung cancer were enrolled and 13-30 peptide-based personalized neoantigens were isolated and identified from the tumor tissue of each patient. Simultaneously, the selected neoantigens were pulsed to DCs induced to differentiate from each patient’s PBMCs. The personalized neoantigen-pulsed DC vaccine was shown to be well tolerated and to induce specific T-cell immune responses, with an objective response rate of 25% and a disease control rate of 75%. Specifically, patients with lung cancer that metastasized to the bone, pelvis, and adrenal and inferior vena cava lymph nodes received personalized neoantigen-pulsed DC vaccine therapy after failure of three lines of therapy, including the anti-PD-1 antibody therapy. After five doses of this vaccine, the metastatic lymph nodes and adrenal and pelvic lesions shrank, indicating a favorable therapeutic effect of this vaccine (51).

We conducted a phase I study on the safety and feasibility of a DC vaccine loaded with synthetic short peptides of neoantigens, identified by whole exome sequencing and RNA sequencing of solid tumors using NGS (jRCTc030190213). Four neoantigens with high binding affinity to MHC class I molecules were selected for each patient and 9-10 mer short peptides were generated. DCs were cultured for 5 days in AIM-V medium supplemented with IL-4 and GM-CSF from peripheral blood monocytes. For maturation, the DCs were cultured for an additional 2 days in AIM-V medium supplemented with GM-CSF, IL-4, TNF-α, prostaglandin E2, and neoantigen peptides as previously described (33). Zoledronic acid was also added as an adjuvant to the DCs during culture (52). In this study, a total of 25 subcutaneous vaccinations were administered to five patients with solid tumors and, except for a Grade 1 fever in one patient with pancreatic cancer, the neoantigen DC vaccine was well tolerated. Of the four patients who completed the five vaccinations of the protocol treatment, positive delayed-type hypersensitivity (DTH) reactions were observed in two patients (Table I). We observed that T cells reacted to the neoantigen peptide pool after administration in one of the three patients who could be studied before and after vaccination (53) (Figure 2). The neoantigen-specific T cell response was detected in only one patient because the activation of CD8+ T cells is essential but the frequency of CD8+ T cell activation is not always high (5). Indel-based mutations may generate more immunogenic neoantigens compared to SNVs (7). Out of four neoantigens selected for each patient, two Indel-based neoantigens were used for N-01, and one Indel-based neoantigen was used for N-02. For the other three patients, only SNV neoantigens were used. We used more Indel-based neoantigens for N-01, which may be another reason why a strong immune response was elicited.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

ELISpot assay in clinical trials of neoantigen-peptide-loaded DC vaccines. ELISpot assay of IFN-γ production before and after vaccination was performed in three patients. IFN-γ expression in PBMCs stimulated by DCs loaded with neoantigen peptide pools (solid line) and by DCs without peptide loading (dotted line) were evaluated pre-vaccination, after 3 and 6 treatments with the vaccine. An immune response specific to the neoantigen peptide pool was observed after vaccination in DTH-positive pancreatic cancer patients (N-01) (p<0.05, Student t-test). ELISpot: Enzyme-linked immunospot; IFN: interferon; DC: dendritic cell; PBMC: peripheral blood mononuclear cell; DTH: delayed-type hypersensitivity.

Neoantigen-vaccine-based combined immunotherapy. In a clinical trial of NEO-PV-01, an anti-PD-1 antibody and neoantigen personalized vaccine, in 82 patients with advanced melanoma, non-small cell lung cancer, or bladder cancer, the combination therapy was safe and well tolerated. CD4+ and CD8+ T cell responses exhibiting a neoantigen-specific and cytotoxic phenotype were observed in all patients after neoantigen vaccination. In addition, after the vaccination, epitope spreading was observed in response to a neoantigen not included in this vaccine (54). Neoantigen-pulsed DC (Neo-MoDC) vaccines in combination with an ICI have been reported for metastatic gastric cancer. Neo-MoDC vaccines alone have the potential to induce neoantigen-specific CD4+ and CD8+ T cell responses, but combination therapy resulted in a stronger immune response and tumor regression (55).

Thus, these findings hold promise for immunotherapies that combine neoantigen-based DC vaccines with other therapeutic strategies to induce a highly patient-specific immune response, leading to a more potent attack against cancer.

Completed or ongoing clinical trials of neoantigen-targeted DC vaccines are shown in Table II. In many studies, neoantigen peptides are loaded into DCs to develop personalized vaccines; the NCT05886439 study is of interest because it is one of the few clinical trials on mRNA-based neoantigen DC vaccines. In addition, the development of combination therapies is being accelerated through studies combining chemotherapy, radiation therapy, and ICI therapy, as well as other therapies, such as effector cell therapy or microwave ablation. Overall, DC vaccines and neoantigen-targeted therapies are making great strides in the field of cancer immunotherapy, but continued comprehensive research and innovation are needed to maximize their potential.

Future Challenges and Prospects

Personalized DC vaccines have shown efficacy in clinical applications, but several research issues must be solved. Although neoantigens are screened for mutations only in the exon region that accounts for 2% of the entire human genome, 99% of tumor-specific mutations occur in noncoding regions of a gene. In leukemia and lung cancer patients, an abundance of both mutated and nonmutated antigens from noncoding regions was discovered using mass spectrometry and RNA sequencing (56). Neoantigens are generated primarily from mutations in coding regions of the genome, but noncoding regions could also generate neoantigens, such as through promoter mutations, intron retention, non-coding RNAs, or mutation-induced novel open reading frames (13). Mass spectrometry has identified noncoding regions that generate large numbers of nonmutated tumor-specific antigens, most of which are derived from epigenetic modifications in atypical translational events (56, 57). Additionally, these tumor-specific antigens in several noncoding regions have been identified as targets for TIL therapy (56). These tumor-specific antigens in noncoding regions are more pervasive than neoantigens in coding regions and may be shared among tumor patients (56, 58). Not only with the personalized neoantigens but also with the shared antigens derived from epigenetic modifications of noncoding regions, future vaccines have the potential to become both personalized and universal cancer therapies. Tumor neoantigen sequencing and screening are complex and time-consuming processes, but in the future, the time and cost of these processes can be significantly reduced. In addition, the production of personalized neoantigen-pulsed DC vaccines requires the individual culture of DCs recovered from each patient by apheresis, and improvements in DC maturation technology during the DC ex vivo culture process may allow for the development of personalized DC vaccines at lower costs (59).