Epstein-Barr virus (EBV) is a ubiquitous gammaherpesvirus that persistently infects more than 90% of the world's adult population. EBV infection primarily targets B cells and epithelial cells. Although infection with EBV is usually asymptomatic, the development of symptomatic disease has been associated with a delayed primary infection leading to infectious mononucleosis in adolescents and young adults, and with various EBV-associated malignancies. The virus also causes various opportunistic diseases in immunocompromised hosts.

In apparent immunocompetent hosts, EBV can also induce chronic disease with prolonged infectious mononucleosis-like symptoms and a sustained EBV DNA load in the peripheral blood. Although rare, this disease has been called chronic active EBV disease or chronic active EBV infection; both are abbreviated as CAEBV (1, 2). Interestingly, while EBV-infected B-cell proliferation has been observed in most patients with CAEBV in Western countries, the virus is found mainly in T or NK cells proliferating clonally in East Asian countries, including Japan (3, 4).

In addition to CAEBV, there is a group of rare diseases that develop in the apparent absence of immunodeficiency and are characterized by ectopic infection of EBV in T or NK lymphocytes. They are EBV-associated hemophagocytic lymphohistiocytosis (HLH), hydroa vacciniforme (HV)-like lymphoproliferative disease, and severe mosquito bite allergy (SMBA). These diseases have a particularly high incidence in East Asia and Latin America and are collectively called EBV-associated T/NK-cell lymphoproliferative diseases (EBV-T/NK-LPDs) (5, 6). However, the EBV-T/NK-LPDs concept has not been fully established. In fact, there are several unanswered questions about EBV-T/NK-LPDs. For instance, it is not known how EBV infects T and NK cells, or how they proliferate. It is also not known why EBV-infected T/NK cells are not removed by host immune responses, or whether they are immunodeficient or malignant diseases. Moreover, the scientific community cannot yet explain why the incidence rates of EBV-T/NK-LPDs are uneven across geographic territories, or whether there are any genetic predisposing factors. Finally, the ideal timing and protocol for hematopoietic stem cell transplantation for CAEBV remain to be determined.

This review investigates both original works and secondary sources on EBV-T/NK-LPDs. It is intended to stimulate discussion on the enigmatic pathogenesis and disease concept of EBV-T/NK-LPDs, to facilitate, eventually, the development of a standard therapy.

EBV preferentially infects B cells via the CD21 cell surface protein, and is associated with a variety of diseases of B-cell origin. These include infectious mononucleosis, Burkitt lymphoma, diffuse large B-cell lymphoma, and Hodgkin lymphoma (7). EBV nucleic acids are detected in 25~50% of Hodgkin lymphoma in the USA and Europe (8, 9). EBV causes immunodeficiency-associated lymphoproliferative diseases, such as post-transplant lymphoproliferative disorders and lymphomas associated with HIV infection (7). EBV can infect epithelial cells and is associated with nasopharyngeal carcinoma and gastric cancer (7). EBV is also associated with diseases of T- and NK-cell origin, such as extranodal NK/T-cell lymphoma-nasal type (ENKTL), aggressive NK-cell leukemia (ANKL), and EBV-T/NK-LPDs as already mentioned. However, the specific T or NK cell receptors that contribute to the disease state are not known.

According to the 2017 World Health Organization Classification of Tumors of Haematopoietic and Lymphoid Tissues, four diseases are classified as EBV-positive T-cell and NK-cell LPDs of childhood: systemic EBV-positive T-cell lymphoma of childhood, the systemic form of CAEBV of the T/NK-cell type, HV-like LPD, and SMBA (10). ENKTL and ANKL are also described in the text and are classified under the umbrella of EBV-positive T/NK-cell proliferation. Since EBV-associated HLH is non-neoplastic, it is not listed as an EBV-positive T-cell and NK-cell lymphoproliferative disease of childhood in this classification. It is, however, important to note that EBV-associated HLH is also characterized by EBV-positive T/NK-cell lymphoproliferation. Furthermore, EBV-associated HLH is an intractable, potentially fatal disease (11). EBV-T/NK-LPD that are described and classified in the 2017 WHO lymphoma classification are shown in Table 1.

The pathogenesis of EBV-T/NK-LPDs is unclear. Some evidence indicates that EBV infects T/NK cells during the primary infection (22), although this does not necessarily lead to EBV-T/NK-LPDs. EBV-infected T or NK cells may proliferate and evade apoptosis with the help of viral oncogenes and may eventually cause EBV-T/NK-LPDs in rare cases (23–25). There are however several unknown mechanisms underlying this process, as summarized in Table 2.

First, the mechanism of T/NK cell infection by EBV remains to be determined. T/NK cells lack CD21 and HLA-DR, both of which are EBV receptors in B cells. Therefore, the exact mechanism of EBV attachment and entry into T/NK cells remains largely unknown (29). Tabiasco et al. reported that NK cells activated by EBV-infected B cells acquire CD21 molecules by synaptic transfer, and these ectopic receptors allow binding of EBV to NK cells (28). EBV may attach to T cells via CD21, which is expressed in immature T cells and common progenitor cells (26, 27, 35). Recently, dual infections of both T-cell and NK-cell lineages with a single clone of EBV have been increasingly reported in patients with CAEBV. EBV may infect lymphoid progenitor cells that express CD21 and have the capacity to differentiate into both T and NK cells, at least in some patients (22).

Second, it is not known whether EBV-T/NK-LPDs are malignant neoplastic diseases like overt leukemia or lymphoma. Based on current knowledge, ENKTL and ANKL appear to be malignant neoplastic diseases. EBV-T/NK-LPDs can be considered as a pre-leukemic or lymphomatous stage in which genetic mutations accumulate and drive the process of lymphomagenesis. Recent comprehensive genetic analyses indicate that DDX3X, TP53, STAT3, and BCOR1 are driver genes that are frequently mutated in patients with ENKTL (36, 37). Interestingly, somatic driver mutations that are seen in ENKTL were found in EBV-infected cells from CAEBV patients. This suggests that EBV-infected T/NK cells evolutionally expand with driver gene mutations, and that mutations in DDX3X are important for the development of overt lymphoma and leukemia in patients with CAEBV (2). Activation-induced cytidine deaminase (AID), which belongs to the APOBEC3 protein family and induces somatic hypermutations and class switch recombination, is a candidate cause of such host gene mutations. Upregulation of AID driven by EBV infection plays an important role in c-myc translocation and tumorigenesis of Burkitt lymphoma and other lymphomas of B-cell origin (38). High expression of AID is seen in EBV-infected T/NK cells from patients with EBV-T/NK-LPDs, and may be associated with driver gene mutations (39).

Finally, although EBV is ubiquitous, it is unknown why only some people develop EBV-T/NK-LPDs in specific areas of the world. While the existence of immunodeficiencies in certain areas has been postulated, mutations in known immune-associated genes are not seen in the majority of cases (33, 34). An association with a specific HLA (A26) was reported, but it did not account for all cases (22, 30). Other explanations suggest that environmental factors are associated with the progression or development of T/NK-cell tumors. In a case-control study, exposure to pesticides and chemical solvents was correlated with ENTKL development (31). Specific EBV strains or variants that have a greater tendency to develop T/NK-cell tumors may exist (32), but there is no direct proof. Thus, the etiologies of EBV-T/NK-LPDs may be multifactorial and they are likely to consist of heterogeneous entities.

The prognosis of EBV-T/NK-LPDs is highly variable. Some patients rapidly develop severe complications, whereas others remain stable without therapeutic interventions. So far, no standard treatment for EBV-associated T/NK cell lymphoma has been established. Additionally, there are no effective antivirals against EBV and no molecular-targeted therapies, unlike rituximab against B cell neoplasms. The outcomes of EBV+ diffuse large B-cell lymphoma or post-transplant lymphoproliferative disorders have improved with the addition of rituximab to conventional chemotherapies (40, 41). EBV-infected T/NK cells are also generally resistant to chemotherapy, due to the expression of the multidrug resistance protein p-glycoprotein (42). To combat difficulties in treatment, hematopoietic stem cell transplantation has been introduced as a curative therapy. Although recent efforts at introducing reduced-intensity conditioning resulted in excellent transplantation outcome (43), the rates of complications associated with transplantation are high (12). It is therefore necessary to develop novel approaches to treat EBV-associated T/NK cell lymphoma.

Proteasome inhibitors, histone deacetylase (HDAC) inhibitors, and Janus kinase (JAK) inhibitors have been tested using in vitro or in vivo xenograft models as possible therapeutic approaches (25, 44–46). Additionally. immune checkpoint inhibitor therapy may be beneficial for patients with EBV-T/NK-LPDs, since programmed death-1 (PD-1) blockade with pembrolizumab has proven effective in relapsed or refractory ENKTL (47). To establish a standard therapy against EBV-T/NK-LPDs, new strategies and extensive further research are needed.

HK and SF contributed to concept development and the writing and review of this manuscript, and gave final approval of the version to be published.