Epstein-Barr virus (EBV), also known as human herpesvirus 4 (HHV-4), is a highly prevalent γ-herpesvirus that infects an overwhelming 90% of the adult population worldwide (1). Since its discovery in 1964 from a Burkitt lymphoma cell line, extensive research has been conducted to investigate its association with cancer (2). In 2020, EBV-associated cancers accounted for an estimated 239,700 to 357,900 new cases and caused 137,900 to 208,700 deaths globally (3). EBV is considered the primary etiological agent associated with multiple epithelial and lymphoid cancers of variable fractions, including nasopharyngeal carcinoma (NPC), gastric carcinoma (GC), Hodgkin lymphoma (HL), Burkitt lymphoma (BL), Diffuse large B-cell lymphoma (DLBCL) and Extranodal NK/T-cell lymphoma, Nasal type (ENKTL-NT). In addition, EBV reactivation can lead to uncontrolled B-cell proliferation in immunocompromised individuals, including post-transplant lymphoproliferative disease (PTLD) in hematopoietic stem cell transplant (HSCT) or solid organ transplant (SOT) recipients and B-cell lymphoma in AIDS patients (4–6).

Despite its ubiquity, most people remain asymptomatic throughout their lifetime, owing to the potent host immune system, especially its cellular immunity, which keeps the virus at bay. However, when cellular immunity is compromised or dysregulated, the virus can replicate unchecked, leading to EBV-associated B-cell malignancies (7). These malignancies express EBV antigens that T cells can specifically target (8). Over the last two decades, the encouraging outcomes of adoptive cell therapy using EBV-specific T cells in treating PTLD have sparked significant research interest. Many clinical trials have been launched to explore their potential application in treating other EBV-related malignancies (9). Recent studies find that EBV latent membrane protein 1 (LMP1), upon ectopic expression in EBV-unrelated cancers, can upregulate TAAs and induce a robust TAA-specific CD4+ CTL response (10), indicating that beyond its oncogenic implications, EBV also has the potential for therapeutic applications in cancer treatment. This review aims to advance our understanding of the roles of T-cell immunity across both EBV-related and unrelated cancers and provide insights to devise more effective immune-based cancer prevention and treatment strategies.

The transmission of EBV occurs through oral means and involves the infection of epithelial cells of the oropharynx, followed by replication and spread to B cells, which are major sites for EBV infection in humans. While EBV predominantly targets B lymphocytes and epithelial cells, it can sporadically infect other human cell types, including T cells and natural killer cells, albeit infrequently (11–13). EBV life cycle is complex and is composed of latent and lytic infections. Only nine proteins contributing to B cell transformation and tumorigenesis are expressed during latent infection. These include six EBV nuclear antigens (EBNA-1, -2, -3A, -3B, -3C, and -LP) and three latent membrane proteins (LMP-1, -2A, and -2B). The latent cycle can be subdivided into four patterns, namely latency III, II, I, and 0, characterized by gradually restricted viral gene expression patterns to evade immune surveillance. Ultimately, EBV establishes persistent residence in memory B cells, characterized by the absence of viral antigen expression (latency 0), thereby evading T-cell recognition and acting as a viral reservoir. The latent-lytic switch is a particularly significant event in the EBV life cycle, but its mechanism remains elusive. EBV can transition to the lytic cycle periodically, resulting in viral replication, shedding, and subsequent transmission (8, 11, 12, 14).

During lytic infection, EBV expresses more than 80 lytic proteins that facilitate the generation of new viral particles (8). The viral lytic cycle is divided into three temporal and functional stages: immediate early (IE), early (E), and late (L). IE gene products are transcription factors in charge of turning on the cascade of expression of lytic genes. Among these proteins, the immediate early proteins BZLF-1 and BRLF-1 act as triggers of the EBV lytic cycle (15). E genes encode enzymes with DNA replication function, and L genes are mostly viral structural proteins.

Several lytic genes are somewhat expressed during latent states. For instance, BHRF1, commonly associated with the virus lytic cycle, remains constitutively expressed as a latent protein in vitro within growth-transformed cells and might contribute to virus-associated lymphomagenesis in Wp-restricted BL (16). Additionally, BALF1, expressed with early kinetics during the lytic cycle, is found in latently infected epithelial and B cells (15). While dispensable for lytic replication and B cell transformation, BALF1 might facilitate efficient transformation, potentially in vivo (15).

Under specific circumstances (17), such as immunosuppression like HIV or immunosuppressive therapy (18), concurrent infections such as CMV, HPV, or coronavirus (19, 20), disruptions in cellular equilibrium like hypoxia (21), or psychological stressors like familial and socio-economic instability (22), EBV can switch from latency to lytic infection, termed viral reactivation, contributing to the dissemination of the virus and its potential to cause various diseases and complications.

In EBV-associated cancers, latent EBV proteins are crucial for tumor pathogenesis, and their expression can classify tumors into distinct categories ( Figure 1 ). In type III latency cancers, cells infected with EBV express a full array of latent proteins, including six EBV nuclear antigens (EBNA1, 2, 3A, 3B, 3C, LP), two latent membrane proteins (LMP1, 2), BamH1-A right frame 1 (BARF1), several small noncoding RNAs, various micro-RNAs, and EBV-encoded small RNAs. All EBNA3 family proteins are highly immunogenic and can be effectively targeted and cleared by T cells in immunocompetent individuals (8, 23, 24). Consequently, type III latency malignancies can primarily be seen in innate or acquired immunodeficient individuals, such as PTLD of HSCT or SOT recipients and B-cell lymphoma in AIDS patients. Type III latency can also be seen in EBV-transformed B cell lymphoblastoid cell lines (LCLs) cultured in vitro.

Type II latency tumors mainly include NPC, GC, some cases of HL, and NKT. These tumors express EBNA1, LMP1, LMP2, and BARF1 and have intermediate immunogenicity.

Type I latency, marked by sole EBNA1 expression, is seen in BL and exhibits constrained immunogenicity.

Apart from latent antigens, some lytic cycle transcripts are also found in certain tumors, which encode molecules known to contribute to tumor growth (25). Among these transcripts, BZLF1 and BRLF1 are the IE transcription factors that master-regulate EBV reactivation/lytic expression. Notably, the expression of some of the immediate early genes, such as BZLF1, in the absence of other lytic genes, particularly those encoding late structural proteins, thereby precluding the formation of infectious viral particles, is termed the abortive lytic cycle. Specifically, it is known as the pre-latent abortive lytic cycle when it occurs just after infection. The abortive lytic cycle has been well-documented in pre-latent cells (26–30) and established tumors (31–34). Furthermore, evidence derived from mouse models (35, 36) supports the notion that the abortive lytic cycle facilitates cell-to-cell viral dissemination and contributes to viral-induced tumorigenesis.

Choi et al. (10) demonstrated in a mouse model that the expression of the EBV signaling protein LMP1 in B cells induces T-cell responses against multiple TAAs. LMP1 signaling enhances the presentation of TAAs on B cells and upregulates the expression of costimulatory ligands CD70 and OX40L, leading to the activation of potent cytotoxic CD4+ and CD8+ T-cell responses against LMP1 (EBV)-transformed B cells ( Figure 2 ). Furthermore, through the ectopic expression of LMP1 on patient-derived tumor B cells to prime T cells, autologous cytotoxic CD4+ T cells can be expanded to target a wide range of endogenous tumor antigens, including TAAs and neoantigens. This innovative approach holds great promise for treating B-cell malignancies and augmenting immune-mediated protection against EBV-unrelated cancers by targeting shared TAAs (96).

Several independent studies have also reported a nonviral, cellular antigen-specific component in the human CD4+ T cell response upon EBV-transformed LCL stimulation in vitro (39, 97). However, these cellular antigens have not been identified and their classification as TAAs remains to be established. Furthermore, clinical studies have observed the detection of T cells specific for nonviral TAAs in the peripheral blood following cytotoxic T lymphocyte (CTL) infusion, which is associated with clinical responses (85). Nevertheless, whether these T cells arise through epitope spreading or are derived from the therapeutic T cells through LCL stimulation is unclear. Therefore, further investigations are needed to identify TAAs expressed by EBV-infected or transformed B cells and to determine their recognition by T cells in individuals with EBV infection (96).

In addition to B cells, whether LMP1 or other EBV antigens can induce the upregulation of TAAs in epithelial cells has yet to be examined. Furthermore, the exact roles of MHC II molecules in cancer remain subject to debate and investigation. Accumulating evidence indicates that tumor-specific MHC II expression is linked to positive outcomes in many cancer types (98) (e.g., breast cancer (99), colon cancer (100), melanoma (101)). However, an opposing functional aspect of MHC II has also emerged. In HLA-DR+ melanoma, MHC II lessens CD8+ T cell activity by inducing LAG3+ and FCRL6+ TILs (102) or recruiting CD4+ T cells to the tumor (103). In the TC-1 mouse model of HPV-related carcinoma, the absence of MHC II molecules promotes CD8+ T cell infiltration and activation, curbing tumor growth (104). Moreover, a recent study examining NPC using single-cell transcriptomics has revealed a dual epithelial-immune feature of tumor cells, characterized by the expression of immune-related genes, including MHC II-coding genes (78), which relates to poor prognosis. This distinct trait also links to CD8+ T cell exhaustion and a suppressed tumor environment (78).

Humans can experience common viral infections like CMV, EBV, and influenza. Once recovered, antiviral memory T cells are retained throughout the body to sense reinfection or recrudescence (105, 106) and are endowed with the capacity for rapid response, sustained vigilance, and cytotoxic prowess (107). Although such virus-specific T cells are abundant within tumors, they may not target tumor cells and are therefore regarded as “bystander-T cells” (108). However, emerging evidence suggests that these virus-specific T cells can still be harnessed for cancer immunotherapy (107, 109–111).

One strategy involves antibody-mediated delivery of viral epitopes to tumors (110, 111), achieved by conjugating virus-derived epitopes with tumor-targeting antibodies. These antibodies bind to specific tumor cell antigens and release immunogenic virus epitopes when cleaved by tumor-specific proteases. The released peptide then binds to free HLA class I molecules at the tumor cell’s surface and can be targeted for destruction by circulating virus-specific CTLs (110, 111).

Another strategy employs viral peptides to mimic a viral reinfection event in memory T cells. Memory T cells can execute a ‘sensing and alarm’ function upon antigen re-exposure (112), and this form of immunotherapy is termed peptide alarm therapy (PAT) (109). Reactivating virus-specific memory T cells through intratumoral delivery of adjuvant-free virus-derived peptide triggers local immune activation. This delivery translates to antineoplastic effects, which lead to a significant tumor reduction of tumor growth in mouse models of melanoma (107) and improved survival in a murine glioblastoma model (109). This approach can reactivate and attract T-cell infiltration into the tumor and transform the immunosuppressive tumor microenvironment into immune-active sites.

The preceding review primarily focuses on αβ T cells, but it is important to note the unique features of γδ T cells that make them appealing in various cancer settings. These features include tissue tropisms, MHC-independent antigen presentation, antitumor activity regardless of neoantigen burden (143), and a combination of T and natural killer cell properties (144–146). In humans, γδ T cells can be categorized into Vδ1+ and Vδ2+ cells, with distinct distributions in mucosal tissues and blood/lymphoid organs, respectively. They play a crucial role in antiviral immune responses in cytomegalovirus (147–150). Emerging evidence suggests that γδ T cells also play a role in primary EBV infection and EBV-associated cancers.

During primary EBV infection, there is an observed increase in the frequency of γδ T cells in the blood of patients with infectious mononucleosis (IM) (151–153). Pediatric patients have a bimodal innate response to primary EBV infection (154), influenced by a dimorphism in TCRγ-chain repertoires (155). Altered γδ T cells have also been observed in patients with EBV-associated malignancies, such as NPC, where the impaired functional capacity of γδ T cells is observed despite an unchanged frequency (156, 157). In a case involving a cord blood transplant recipient with elevated EBV viremia, the absence of detectable αβ T cells was compensated by expansions of cytotoxic Vδ1+ γδ T cells, resulting in no signs of lymphoproliferative disorder (158). Moreover, early recovery of Vδ2+ T cells has been identified as an independent protective factor against EBV reactivation in recipients of allo-HSCT (159). Interventions that induce early reconstitution of autologous γδ T cells could hold therapeutic benefits. αβ TCR graft depletion (160, 161) has demonstrated efficacy in reducing GVHD by facilitating rapid immune reconstitution of NK cells and γδ T cells (162–164). Additionally, reducing immunosuppressants has led to enhanced recovery of Vδ2+ T cells and decreased risk of EBV-associated lymphoproliferative disorders in HSCT recipients (159). Notably, long-term persistence of donor-derived Vδ1+ T cell clones has been detected in recipients’ blood even a decade post-HSCT, with these cells exhibiting expandability in vitro and cytotoxicity against autologous EBV-LCL (165).

While extensive research and clinical trials have explored the therapeutic potential of γδ T cells in managing solid tumors and hematopoietic malignancies (166–168), only a limited number of studies have investigated their efficacy in EBV-associated cancers using murine models ( Table 1 ). Adoptive transfer of anti-γδ TCR antibody-expanded γδ T cells to Daudi lymphoma-bearing nude mice significantly prolonged their survival time (130). In addition, the adoptive transfer of pamidronate-expanded Vγ9Vδ2-T cells prevented and inhibited EBV-LPD in mouse models (131). Moreover, co-administration of Vδ2+ T cells and the EBNA1-targeting peptide L2P4 enhanced γδ T cell cytotoxicity against NPC in immunodeficient mouse models (132). Additionally, exosomes derived from Vδ2+ T cells exhibited the ability to eliminate EBV-associated tumor cells (133), and when combined with radiotherapy, γδ-T-Exos demonstrated efficacy in effectively treating NPC by eradicating radioresistant cells (134).

Thus, γδ T cells represent an essential component of cellular immunity in regulating primary EBV infection and hold promise in combating EBV-associated malignancies.

Cellular immunity is pivotal in maintaining the delicate equilibrium between the host and EBV. Despite EBV’s high prevalence, affecting a significant portion of the global population, most individuals remain asymptomatic throughout their lives, highlighting the critical role of effective immune control. However, EBV-associated malignancies primarily occur in individuals with apparently intact immune function. This raises intriguing questions about the mechanisms and stages at which these tumors manage to evade the surveillance of virus-specific T cells.

EBV-associated malignancies express distinct EBV latent antigens, triggering diverse T-cell responses while also employing a range of immune evasion mechanisms, rendering a complex interplay with cellular immunity. Encouragingly, promising clinical responses have been observed from adoptive cell transfer of EBV-specific T cells targeting latent antigens. Recent investigations into early lytic EBV antigens in tumorigenesis provide additional potential targets for therapeutic interventions. Additionally, TCR transgenic therapy offers the possibility of redirecting T cells to recognize EBV antigens and the involvement of γδ T cells also merits consideration in EBV-associated diseases.

In cancers not associated with EBV, there usually exists an abundance of EBV-specific memory T cells, which can be leveraged to either activate the immunosuppressive tumor microenvironment or re-directed to target tumor cells. In addition, EBV can activate TAA-specific T-cell responses. These further broaden our understanding of this oncogenic virus and its implications for the fields of cancer biology and therapy. In this regard, a pivotal research goal is to attain a comprehensive grasp of the intricate interplay between cellular immunity and the virus. By harnessing the inherent capabilities of T-cell immunity, we can advance toward more precise and effective interventions in the treatment of EBV-associated and other cancers.

QZ prepared the manuscript draft, MX provided valuable comments to the text and figures, read and approved the final manuscript. All authors contributed to the article and approved the submitted version.