2.1 Neoantigens are generated by somatic mutations and viral oncogene integrations

Gene mutations caused by genetic instability during carcinogenesis occur in noncoding or coding regions. The amino acid changes caused by these mutations lead to the production of proteins that do not exist in normal cells. These aberrant proteins, which can be recognized by immune cells, are so-called tumor antigens [1] and lead to the immune system's attack on cancer cells [2].

Tumor antigens are generally categorized into tumor-associated antigens (TAAs) and tumor-specific antigens (TSAs). The conventional TAAs are different from TSAs, namely neoantigens. Several decades of tumor antigen studies have led to the discovery of a wide range of TAAs, including those that are implicated in tissue differentiation or typically overexpressed in cancer cells but not in normal tissues (other than fetal tissues or immune-privileged organs) [3]. TAAs such as carcinoembryonic antigen (CEA) and melanoma-associated antigen are mostly overexpressed differentiation antigens or cancer-testis antigens that appear in a subset of patients [4]. Mammaglobin-A is normally expressed in the mammary gland but overexpressed in breast cancer; prostate-specific antigen is expressed in the healthy prostate gland near prostate cancer; and melanoma antigen recognized by T cells 1 (MART1), melanocyte protein pre-melanosome protein (PMEL), and tyrosinase are expressed in normal melanocytes but not in melanoma cells. These are classic examples of this type of TAA [5]. However, because TAAs are also expressed in normal tissues, they are subject to both central thymic tolerance and peripheral tolerance, compromising their effectiveness as therapeutic targets [6]. In addition, targeting TAAs may lead to autoimmune toxicity [7].

Nonsynonymous somatic mutations lead to unique epitopes of self-antigens that are processed and presented by the major histocompatibility complex (MHC) and then recognized by T lymphocytes [7]. Insertions or deletions (indels), point mutations [8], and gene fusions [9] are typical mutations that lead to neoantigens in various cancers [10]. Because tumor neoantigens are generated from somatic mutations, they are not only highly tumor specific but also immunogenic because of the lack of central tolerance during their generation [3]. Cervical cancer or head and neck squamous carcinoma associated with human papillomavirus (HPV), hepatocellular carcinoma associated with hepatitis B virus, and Kaposi sarcoma associated with human herpes virus 8 have all been related to oncogenic viral gene-encoded antigens [11]. Because of their origin, truncated or fusion oncoviral proteins are regarded as an alternate class of neoantigens in cancers driven by viruses or integrated with driver viral oncogenes [12]. These viral neoantigens are particularly ideal targets for cancer vaccines because they are foreign to the host (and hence not subject to central tolerance) and expressed solely by cancer cells (and thus unique in the tumor). Vaccinations targeting these antigens have shown clinical benefit in the prevention and therapeutic strategies of HPV-related malignancies [11, 13].

Currently available immunotherapy strategies are used to treat hematological malignancies and several types of solid tumors. Depending on whether the immunotherapies are passive or active, immunotherapy employs tumor reactive immune cells or antitumor antibodies to destroy the cancer cells. Because recognition of cancer cells by the immune system is an important antitumor mechanism, it becomes more critical to identify the tumor neoantigen before precise medical treatment is administered. TSA neoantigens are thus required for the development of effective immunotherapies, such as immune-checkpoint blockade therapies, customized personal cancer vaccines, and ACT therapies of antigen-specific T cells [14-16].

2.2 Neoantigens: TSAs recognized by CD4⁺ and CD8⁺ T cells

Neoantigens are the targets of a successful tumor-specific immune reaction. In addition, tumors with a high mutational load are thought to contain more immunogenic neoantigen possibilities for eliciting CD4+ and CD8+ T-cell responses.

Tumor antigen neoepitopes, short peptide molecules approximately 8–18 amino acids in length, play an important role in both adaptive and active immunotherapies [17]. The mutated neoantigen peptides are first degraded by the proteasome and then transported to the endoplasmic reticulum (ER) by the peptide transporter protein transporter associated with antigen processing (TAP). In the ER, they bind to MHC to form the peptide-MHC complex (pMHC) to be presented on the surface of APCs. These neoantigen epitopes presented via MHC class I and II molecules then activate CD8+ and CD4+ T cells through the T-cell receptor–peptide-MHC complex (TCR-pMHC), respectively [18].

Understanding of peptide-MHC and T-cell receptor interactions is critical for building peptide-based cancer vaccines. Short peptides (usually nine amino acids) attach to MHC molecules directly; these are more likely to establish immune tolerance and are easily degraded [19]. Longer peptides (approximately 30-mers) may be stronger immune stimulators as they are internalized and digested by APCs and presented by the MHC molecule for CD4+ and CD8+ T cells to generate better long-term memory [19].

Neoantigens are widely recognized as critical targets for effective antitumor immunity [20]. First, a larger tumor neoantigen load has been linked to stronger T-cell responses and better clinical outcomes in multiple studies. Neoantigen frequency per tumor type was proven to be positively linked with gene expression patterns of T-cell cytolytic ability from RNA-sequencing (RNA-seq) data from 18 solid tumor types of The Cancer Genome Atlas (TCGA) database [21].

The significantly larger load of predicted immunogenic epitopes was connected to improved patient survival in research involving 515 tumors from 6 distinct histologies from TCGA [22]. From these studies, whole-exon sequencing of 619 colorectal tumors revealed that a high neoantigen burden is positively connected with increased tumor infiltrating lymphocyte (TIL) infiltration and longer survival [23]. The link between tumor neoantigen load and TIL number in other malignancies, like endometrial cancer, has also been demonstrated [24].

Second, in the presence of effective anticancer immunity, neoantigen-specific T-cell populations grow abundantly. This has been seen in melanoma patients with better clinical responses to ipilimumab (anti-CTLA4 antibody) [25, 26] and in NSCLC patients who received pembrolizumab (anti-PD1 antibody) [27]. Furthermore, tumor regression induced by CD4+ and CD8+ T cells in patients receiving adoptive TIL transfer has been shown to exhibit neoantigen specificity [25, 28-30].

Third, a preclinical animal model study and clinical trials showed that neoantigen-specific T lymphocytes are cytolytic for tumor cells containing altered peptides, resulting in regression of tumor bulk. The epitopes recognized by CD8 T lymphocytes in rejected tumors were identified as neoantigens in both a transplantable chemically generated sarcoma model [31] and an induced sarcoma model with transfected immunodominant antigens [32].

T-cell responses generated by neoantigen peptide vaccination stimulated anticancer activity in a melanoma model and a transplantable colon cancer mouse model in both preventive and therapeutic situations [33-35]. A neoantigen long peptide vaccination (capable of triggering both CD4+ and CD8+ T-cell responses) caused tumor rejection equivalent to that generated by ICI treatment in another chemical-induced sarcoma mouse model [15]. In mice, neoantigen vaccines administered as poly-neoepitope mRNA, containing both MHC class I/II-restricted neoepitopes, generated powerful tumor-specific immune responses and rejection of melanoma and colon cancer xenografts [36]. Other preclinical animal models for evaluation of therapeutic neoantigens have also shown promise. Adoptively transplanted neoepitope-specific CD4+ T cells promoted tumor regression in a cholangiocarcinoma patient, giving direct proof for neoantigen-specific T-cell anticancer efficacy [14]. Thus, neoantigens are ideal targets for therapeutic cancer vaccines and T-cell-based ACT immunotherapy.

2.3 Neoantigens and tumor heterogeneity

During tumorigenesis, acquired mutations lead to the creation of neoantigens that represent intratumoral heterogeneity (ITH). Tumors contain a high level of genomic heterogeneity, which has significant implications for cancer therapeutic effectiveness, including ICI or CAR-T-cell immunotherapy [37]. Because of this heterogeneity, the bulk of a tumor might have a diverse collection of cells harboring distinct molecular signatures with differential levels of sensitivity to treatment. This heterogeneity might result in a nonuniform distribution of genetically distinct tumor cell subclones within primary or metastatic sites (spatial heterogeneity) or temporal variations in the molecular expression of cancer cells (temporal heterogeneity) [37].

The capacity to identify tumor-specific neoantigens in specific patients is critical for navigating T cells to eradicate tumors by vaccination. However, tumors with high-level ITH might predispose patients to inferior clinical outcomes. Under therapeutic pressure, treatment resistance can emerge from the expansion of pre-existing subclonal populations or the evolution of drug-tolerant cells. Through integrated analysis of ITH and neoantigen burden, researchers identified a relationship between clonal neoantigen burden and overall survival in lung adenocarcinomas [38]. T cells recognizing clonal neoantigens were detectable in patients with durable clinical benefit. Cytotoxic chemotherapy-induced subclonal neoantigens, contributing to an increased mutational load, were enriched in certain poor responders. These findings demonstrate a phenomenon that homogeneous tumors (neoantigen ITH ≤ 1%) with higher tumor neoantigen burden are associated with longer overall survival compared with heterogeneous tumors. Thus, subclonal neoantigens induced by chemotherapy may cause immune tolerance and impair the immune response against cancer cells [38]. As a result, monitoring of tumor heterogeneity using tumor-specific antigens and the prediction of immunotherapy treatment effectiveness are critical. Patients with tumors with a high TNB and low ITH are more likely to respond to immunotherapy [39]. It is thus likely that a single vaccination to target many neoantigens is required to limit the risk of immune evasion and successfully eliminate the entire tumor [40].

3.1 Beginning with cDNA library screening for neoantigen isolation

In the early 20th century, research revealed that the immune system recognizes and eliminates tumor cells. However, the nature of antigens that trigger the antitumor immune response was unclear. In 1988, De Plaen and colleagues identified the first neoantigen recognized by T cells in a mouse tumor model using cDNA library screening [41, 42]. The researchers discovered that even if there was only one nucleotide difference between the normal and tumor gene, this mutation could still cause an amino acid change. Following this study, multiple neoantigens originating from somatic mutations were discovered in melanoma and renal cell carcinoma in humans [42].

Most of the unique neoantigens were discovered through cDNA library screening in the last two decades, published in studies before 2013. In this approach, MHC molecules and cDNA libraries were overexpressed in cell lines and then cocultured with T cells to identify antigens that activate T cells, detected by cytokine release or 4-1BB upregulation [43]. Bacteriophage−eukaryotic cell interaction provides the biological foundation of the phage display technique [44]. Phage display has also enabled the development of novel therapeutic medicines that target specific cancer mutations. Proteins encoded by cancer cell-mutated genes, namely neoantigens, can be processed and presented on the tumor cell surface by human leukocyte antigen (HLA) molecules [45].

Neoantigens have primarily been discovered in melanoma using these technologies, most likely because this type of tumor has a relatively high mutation rate. Neoepitopes have also been found in a variety of malignancies, including lung and renal cancer [43]. Although frameshift deletion and insertions were shown to generate neoepitopes, most neoantigens are encoded by point-mutated gene products. Some of the altered gene products identified by T cells appear to be driver mutation products, meaning they play a key role in carcinogenesis. These include CDK4, β-catenin (CTNNB1) and Caspase-8 (CASP8) proteins [46-48]. In contrast, the traditional cDNA library screening method is time-consuming and low-throughput and unable to detect some altered antigens produced from GC-rich transcripts and low-expression transcripts [43, 49].

3.2 Improving neoantigen prediction in silico with mass spectrometry

Recent technological advances in next-generation sequencing, bioinformatics, single-cell proteomics, and transcriptomics have enabled the research and identification of personalized neoantigens [50]. A peptide-based screening approach involving whole-exome sequencing (WES) and MHC-peptide binding prediction algorithms has been successful in identifying neoantigens recognized by TILs in melanoma patients [30]. The use of tandem minigenes (TMGs) composed of multiple minigenes that encode polypeptides containing a mutated amino acid residue flanked on their N- and C-termini by 12–13 amino acids were synthesized and used to transfect APCs; these TMGs have resulted in the identification of neoantigens in melanoma, colorectal carcinoma, and cholangiocarcinoma patients and mouse tumor models [14, 36, 51, 52].

Computational algorithms have been applied in the detection and prioritizing of neoantigens related with cancer vaccination research [53-57]. Details of this study have been reviewed thoroughly [3, 58, 59]. Here, we briefly outline the neoantigen prediction methodology, focusing on advances of recent years [20].

In neoantigen prediction methodology, the patient's tumor biopsy specimens and nonmalignant tissue samples (usually peripheral blood mononuclear cells) are acquired for WES, which compares tumor DNA with germline DNA to discover tumor-specific somatic alterations. RNA sequencing further reveals the expression of altered genes and confirms the mutations. Many factors should be considered when predicting tumor neoantigens, such as expression level, mutation variations, peptide prediction, TCR binding force, HLA typing, MHC affinity, pMHC stability, and tumor neoantigen source [60]. Depending on the tumor type, many tumor-specific mutations may be found; however, because of HLA restrictions, not all mutations result in neoepitopes that are recognized by the immune system. Because there are over 16,000 distinct traditional HLA class I alleles (HLA-A, HLA-B, and HLA-C), it is the first necessary for HLA typing for the successful prediction of immunogenic epitopes [61].

Computational techniques are used to estimate MHC I-binding epitopes, and peptides with projected HLA-binding affinity in the moderate-to-strong range (IC50 < 150 nmol/L) are considered to be more likely to elicit CD8+ T-cell responses [62]. Various computational methods have been described for predicting epitopes presented by MHC I [63]. Traditionally, peptide binding affinity data have been used to train these algorithms [64].

So far, epitope prediction algorithms have mostly concentrated on MHC I-binding epitopes, while the MHC II-binding epitope prediction is less developed. This is because of the closed ends of the MHC I peptide-binding groove, which dictate the placement of 8–11 amino acid peptide epitopes for presentation to CD8+ T lymphocytes. In contrast to the MHC I groove, which has open ends, the MHC II groove may recognize longer peptides of various lengths; its binding to the MHC II molecule can be impacted by the elongated peptide flanking the MHC II core, and these peptides can also bind to multiple MHC II molecules [65]. Together, these characteristics of MHC II peptide binding make identifying immunogenic MHC II epitopes more difficult. The epitopes of MHC II, in contrast, have been included into vaccine peptide design in an indirect way. Long peptides (15–30 mers) with expected MHC I-binding CD8+ T-cell epitopes are used in immunization techniques to improve peptide absorption and processing by professional APCs, enhancing T-cell activation [13, 66, 67]. In addition to the presentation of predicted CD8+ T-cell epitopes on MHC I proteins, this long-peptide approach is compatible with the processing and presentation of CD4+ T-cell epitopes by MHC II proteins. Following immunization, CD4+ T-cell responses were preferentially induced above CD8+ T-cell responses in many clinical studies [36, 54-57]. This finding might be because of, at least in part, the specific characteristics of MHC II open-ended peptide-binding groove with the ability to offer a wider range of peptides than the more rigorous MHC I groove [68]. Other immunological variables, such as changes in the stimulation of dendritic cell (DC) subsets that favor CD4+ T-cell activation or the generation of CD8+ T-cell responses by cross-presentation of antigen, might contribute to the dominance of vaccine-mediated CD4+ T-cell responses. Other immunological factors, like shifting into a DC stimulatory subset that favors CD4+ T-cell activation or the development of CD8+ T-cell responses via antigen cross-presentation, might also contribute to the CD4+ T-cell dominant responses mediated by vaccination [69].

Mass spectrometry (MS)-based techniques have also been used with exome sequencing and MHC I-binding prediction algorithms to discover peptides that can be processed and presented by tumor cells. However, MS requires a large volume of tumor tissues and the sample-processing procedures are complicated. Novel prediction models trained on MHC I-binding peptides eluted from single HLA-allele-expressing cell lines and identified by MS revealed that adding antigen-processing information and de novo development of new peptide motifs might increase the accuracy of these algorithms [70].

Improved prediction algorithms derived from peptides eluted from MHC proteins have been developed through MS [71]. The algorithms that were developed using peptides presented by mono-allelic HLA-expressing cell lines can help with antigen processing and presentation procedures of endogenous neoantigens [70, 72]. By increasing the training HLA allele numbers, these MS-based algorithms can be potentially improved for accuracy [73]. Further studies are required to optimize neoantigen expression, presentation processes, and enhancement of immunogenicity [74].

CD4+ T cells can orchestrate antitumor immunity and, via their traditional “helper” functions, play a crucial role in generating and maintaining CD8+ T-cell responses, prompting efforts to generate new epitope prediction tools. In 2019, one study reported a machine learning prediction system, MixMHC2pred, that used an MS-based technique combined with motif deconvolution and annotation training [75]. In addition, in 2020, the NetMHCIIpan MHC II–binding epitope prediction algorithm, which was developed using peptide binding affinity data from the Immune Epitope Database (IEDB), was updated to include the NNAlign MA algorithm, which was trained using MS-derived data on peptides eluted from MHC II proteins [76]. These studies emphasize a current trend on increasing the effectiveness of therapeutic cancer vaccines by strengthening tumor-specific CD4+ T-cell responses.

The Tumor Epitope SeLection Alliance (TESLA), a bioinformatics cooperation including scientists from well-known neoantigen research organizations, was formed to produce a global neoantigen prediction algorithm standard [77]. TESLA brings together 36 top biotechnologies, pharmaceutical, university, and nonprofit research teams including the National Cancer Center, Parker Institute for Cancer Immunotherapy, Memorial Sloan-Kettering Cancer Center, MD Anderson Cancer Center, and more than 30 top neoantigen research institutions. These institutions independently study the open database of tumor sequencing to predict potential neoantigens and candidate peptide ranks. Different predictions are collected, analyzed, and crossmatched to reach a final optimized version. This integration includes aspects of binding affinity, tumor abundance, stability, and peptide identification in addition to antigen presentation. Therefore, better recognition precise rate would be expected [77]. Future advances in prediction algorithms for finding neoantigens arising from splice variants, gene fusions, and translation errors are likely to increase the number of targetable neoantigens, which is particularly important for tumors with low mutation loads [3].

4.1 ICI efficacy is better correlated with tumor neoantigen burden (TNB) than tumor mutational burden (TMB)

Immunotherapy can generally be divided into two categories: active and passive. Active immunotherapy refers to the eradication of cancer cells by stimulating the body's immune system by a cancer vaccine such as peptide vaccine, mRNA vaccine, DNA vaccine, and DC vaccine. Passive immunotherapy refers to the passive acceptance of antibodies, cytokines, and/or transformed immune cells that act directly on a tumor [78]. Cancers use strategies to reduce T-cell responsiveness to avoid immunological elimination. ICIs have revolutionized cancer treatment by reactivating T cells. However, the immune response rate varies widely, and there is a lack of adequate biomarkers to identify responders from nonresponders. As indicated by the NCCN guidelines, the biomarkers of the response to ICI immunotherapy focusing on programmed death-ligand 1 (PD-L1), TMB, and microsatellite instability (MSI) are still not sufficient for immune response prediction.

TMB is defined as the number of mutations present in a megabase of the genomic region by WES or large-scale next-generation sequencing [79-81]. TMB indicates the frequency of tumor mutation, and a larger TMB is more likely to include possible neoantigens, which enhances T-cell recognition and clinically correlates with improved ICI effectiveness. Nonetheless, the significance of TMB in immunotherapy remains unclear, according to various clinical trial results, because only nonsynonymous mutations generate neoantigens [82-84]. Only a minimal number of genetic changes can be appropriately processed, presented on the cell surface MHC complex, and identified by T cells [85].

TNB is quantified by the neoantigens detected per megabase in the genomic area [86, 87]. Studies have shown that TMB and TNB show a positive correlation. This suggests that TNB may be a superior biomarker for neoantigen assessment and immunotherapy compared with TMB [7, 88]. TNB was also correlated to the expression of macrophage M1 polarizing genes such as PD-1/PD-L1, granzyme B, FAS and FAS ligand, interferon-γ (IFNγ) genes and other important immuno-modulatory regulator genes [89]. Several investigations have confirmed the positive correlation between TNB and T lymphocytes infiltrated in the tumor microenvironment, namely TIL cells [89-91]. Clinical trials have also linked higher TNB to better results in patients treated with immunotherapy [92].

It is conceivable that TNB or the quantity of neoantigens will not only replace TMB as a biomarker for successful immunotherapy but also may define the treatment strategy for individualized precision treatment.

4.2 Cellular and molecular neoantigen cancer vaccines

In addition to providing value in ICI therapy efficacy prediction, neoantigens are critical in individualized immunotherapy practices such as cancer vaccines and neoantigen-specific T-cell therapy [93, 94]. Cancer vaccines do not function as traditional vaccines that are used for the prevention of viral infectious diseases. Cancer therapeutic vaccines activate and enhance antigen-specific immune responses against TSAs, such as neoantigens, to possibly trigger T-cell responses for cancer eradication [20]. Table 1 summarizes the current peptide vaccine and cell-based vaccination methods being investigated.

TAA-specific T cells are prone to central and/or peripheral resistance; hence, early therapeutic vaccination approaches based on self-antigens, namely TAAs, were mostly ineffective in eliciting clinically effective anticancer immune responses [105]. TAAs are likely to be detected in normal healthy tissues to some extent, increasing the potential of autoimmune destructive reactions from vaccines [105]. As a result, the lack of tumor selectivity and poor immunogenicity were highlighted as important problems in the development of cancer vaccines in early studies.

The current vaccines predicated on neoantigens rather than traditionally used TAAs have several advantages. First, neoantigens are produced only by tumor cells and thus can elicit truly tumor-specific T-cell responses, thereby preventing severe “off-target” side effects to nonmalignant organs. Second, neoantigens are de novo epitopes generated from somatic mutations, allowing for the circumvention of T-cell central tolerance of self-epitopes and the consequent induction of immune responses against malignancies.

Cancer vaccination platforms are classified as either cellular or molecular (peptide, DNA, or RNA) [106, 107]. Cellular cancer vaccines are generated from autologous patient-derived tumor cells or allogeneic cancer cell line-derived cells [108]. During immune recognition of tumors, DC function as important consumers, processors, and presenters of tumor antigens, which are also used to create cellular vaccines. Better understanding of the peptide−MHC and TCR−peptide−MHC interactions helps to build peptide-based cancer vaccines [1]. TAAs expressed by closed circular DNA plasmids (naked DNA) are coupled with immunomodulatory chemicals to generate tumor-specific responses in classical cancer DNA vaccinations [109]. The advantages of this approach include simplicity, ease of fabrication, and safety; however, because of poor transfection rates into target tumor cells, naked DNA vaccines have limited effectiveness. mRNA vaccines are easily produced in vitro to encode antigen(s), and after internalization, these are translated into proteins that trigger an immune response [110]. mRNA vaccines could carry a large quantity of antigens and costimulatory signals without causing transgenic mutation or gene insertion, and they can be made rapidly and inexpensively; however, the use of mRNA vaccines remains limited because of instability and lack of delivery efficiency [109].

Because most neoantigens are specific to each patient, maximizing the potential of this diverse set of targets necessitates tailored therapy strategies. As a result, personalized neoantigen-based vaccinations provide an extra weapon in the immunotherapy toolkit by boosting tumor-specific immune responses [20]. Several computational methods and machine-learning techniques have been developed to discover cancer mutations in sequence data, to prioritize those that are more likely to be recognized by T cells, and to construct personalized vaccines for each patient [57].

The neoantigen-specific T cells boosted by vaccines show the potential to persist as posttreatment immunological memory and enhance the possibility of long-term disease control of recurrence. The high costs and time delays associated with developing personalized vaccines, as well as the uncertainty over the optimum neoantigen identification platform and a lack of consensus on the best vaccine delivery platform, are all drawbacks of this personalized immunotherapy approach [20]. The next milestone for scientists and researchers is to find new strategies to develop and investigate neoantigen cancer vaccines.

4.3 Neoantigens and ACT therapy

The current precision medicine of cancer, such as immune checkpoint blockade therapy or CAR-T therapy, still shows limitations in efficacy. The extensive clinical antitumor efficacy of ACT and cancer vaccination has significantly fueled the rise of cancer immunotherapy applications. The ability to harness this vast tumor cell-specific repertoire of highly immunogenic antigens for customized cancer vaccines requires the knowledge of an individual's cancer mutanome [10]. Thus, the combination of cancer vaccine with ACT and ICI is a potentially potent strategy to attack the immunosuppressive tumor microenvironment.

T cells, which play an important role in anticancer immunity, boost the immune response by recognizing tumor cells and interacting with MHC-bound cancer-specific peptides. Tumor-infiltrating lymphocytes are an essential therapeutic strategy because neoantigen vaccination alone is insufficient for eliciting an effective immune response [111, 112]. Recent achievements in TIL adoptive transfer therapy field have provided more evidence for enrichment of the arsenal of immunotherapeutic approaches. Various clinical trials showed that autologous tumor-reactive T cells generated from TILs and genetically modified lymphocytes expressing highly active T-cell receptors (TCRs) or CARs have anticancer capability [113].

TNB levels beyond a certain threshold can stimulate T-cell identification and activation, resulting in a rise in TILs and improved immunological responses. Neoantigen-stimulated lymphocyte infiltration and higher levels of proinflammatory cytokines are seen in colorectal cancers with different Mismatch Repair (dMMR). Moreover, TNB alone is also positively related to the tumor's inflammatory microenvironment [114]. Furthermore, lung adenocarcinoma patients with a higher TNB had more infiltrating activated CD4+ and CD8+ T cells, suggesting that the detected mutations might be predictable by higher TNB and T-cell infiltration. TCR clones and the activation of infiltrating T lymphocytes are linked up with high TNB levels [115, 116]. This may be mediated by the increased production of chemokines induced by IFNγ, such as chemokine ligand CXCL9, resulting in the recruitment of cytotoxic T cells or antigen-presenting dendritic cells [117, 118].

CAR-T-cell therapy is intended to target a single antigen, and the clinical effectiveness of this treatment has been confined to patients with B-cell lymphomas, which have uniform tumor cells that express a shared dominant antigen (such as CD19). The lack of a common surface antigen target in solid tumors is a key obstacle for this treatment strategy. CAR-T therapy against conventional TAAs, such as CEA, HER2, GD2, GPC3, and mesothelin, showed limited efficacy in solid tumors [119-121]. However, several clinical trials of CAR-T cell against the newly discovered neoantigen Claudin18.2 (CLDN18.2) showed promising therapeutic potential in solid tumors like gastric cancer, pancreatic cancer, cholangiocarcinoma, ovarian cancer, and lung adenocarcinoma [122, 123]. In the ASCO conference in 2022, the clinical trial of BNT211, a combination of mRNA vaccine with CAR-T targeting Claudin 6, showed a great therapeutic effect, which suggests promise for this treatment strategy [124].

Similarly, despite some encouraging outcomes, the objective response rate of single-agent ICIs is restricted to 30% in most tumor types [125-127]. Another promising direction is TIL adoptive transfer therapy, in which patient-specific T cells are isolated, expanded, and stimulated in vitro. After this complex ex vivo procedure, a defined amount of functional T cells is given back to the patient. By including T helper cells or peptides activating T helper cells, a humoral immune response can be induced [128]. Therapeutic antibodies, adoptive neoantigen-specific TCR-T cells, and multiclonal TIL cell-based adoptive transfer immunotherapies will therefore become strong weapons in the fight against cancer [129].