Introduction

The Epstein–Barr virus (EBV) infects T or NK cells, albeit uncommonly, and can cause benign or severe disease depending on the immunological response of the individual [13]. Nasal-type NK/T cell lymphoma and aggressive NK cell leukemia are prototypes of EBV-associated NK/T cell neoplasms, and the diagnostic criteria are well defined in the WHO classification: Tumors of the Haematopoietic and Lymphoid Tissues [4, 5]. On the other hand, chronic active EBV infections show considerable variations in their biology and pathology when attempting to distinguish between infectious and neoplastic diseases, which lead to difficulties in diagnosis and treatment [6]. Currently, chronic active EBV infection is defined as a systemic illness of greater than 3 or 6 months duration that begins as a primary EBV infection, with histopathological evidence of major organ involvement and increased levels of EBV RNA or proteins in affected tissues or blood [79]. Such a definition of chronic active EBV infection is based mainly on clinical and laboratory findings. Chronic active EBV infections are always accompanied by lymphoproliferation to varying degrees. This syndrome is polyclonal, oligoclonal, or monoclonal in nature and can often transform to an obvious T or NK cell lymphoma or to leukemia [8, 10, 11]. Because chronic active EBV infections show a broad spectrum of lymphoproliferation in tissue biopsies [10], there has been debate regarding the application of the term “chronic active EBV infection” to this spectrum of diseases and on the diagnostic criteria for distinguishing them from overt malignant lymphomas.

Other diseases associated with EBV infection raise similar issues. Systemic T cell lymphoproliferative disease (LPD) of childhood and hydroa vacciniforme (HV)-like T cell lymphoma are described in the new WHO classification [12]. Systemic EBV-positive T cell LPD is defined as a clonal proliferation of EBV-infected T cells with an activated CD8-positive cytotoxic phenotype [12, 13]. It can occur shortly after a primary acute EBV infection or in the setting of chronic active EBV infections. Clinical syndromes associated with a cytokine storm are characteristic of this disease and include fulminant hemophagocytic syndrome, pancytopenia, and multiorgan failure [1416]. Proliferating T cells in this disease should be clonal by definition; however, there have been reports of cases that are polyclonal for T cell receptor (TCR) gene rearrangement in which the proliferating cells might be NK cells or polyclonal T cells [12, 17, 18].

Hydroa vacciniforme-like T cell lymphomas have been named as such because gross findings of the lesion are similar to those for HV, a self-limited EBV-associated vesiculobullous disease healed with central umbilication [19]. Most patients with HV regress spontaneously, but some develop systemic or cutaneous T or NK cell lymphoma or leukemia [2023]. The relationship between HV and HV-like T cell lymphoma remains to be clarified and there are no convincing criteria to distinguish between these two conditions. On the other hand, mosquito bite hypersensitivity is a cutaneous manifestation of chronic EBV infection characterized by intense local skin symptoms including erythema, bullae, ulcers, and scar formation and by systemic symptoms such as fever, lymphadenopathy, and liver dysfunction after mosquito bites.

As described above, T and NK cell proliferative diseases associated with acute and chronic EBV infections have several issues regarding the diagnostic criteria and terminology. The name of the disease must be finely tuned to reflect its exact biology and diagnostic criteria must be redefined.

The 4th Asian Hematopathology Symposium and consensus meeting

Because the T or NK cell proliferative diseases associated with acute or chronic EBV infection are uncommon even in Asia, it is difficult to study a sufficient number of cases to obtain comprehensive knowledge for each form. Therefore, expert hematopathologists from Japan, Korea, Singapore, and Taiwan met in Seoul during the 4th Asian Hematopathology Workshop held on January 29, 2012 with the aim of obtaining a consensus viewpoint for the issues raised. The goal of the meeting was to collect opinions on the terminology and diagnostic criteria for acute and chronic EBV infection-related diseases. In the workshop, expert pathologists presented their work on several topics related to T/NK LPDs associated with chronic and acute EBV infection (Table 1).

Table 1 Topics presented in the 4th Asian Hematopathology Symposium

Here, the deliberations of the expert panel are presented in response to four key questions.

Systemic EBV-positive T cell LPD of childhood

Question 1: Do you agree with the current diagnostic criteria for systemic EBV-associated T cell LPD of childhood?

None of the experts agreed with the current diagnostic criteria and together proposed the following guidelines.

  1. 1.

    It should include NK cell cases.

  2. 2.

    It should include polyclonal, oligoclonal, and monoclonal cases.

  3. 3.

    The terminology of “childhood” in the current WHO classification might be better changed to “childhood-type” because a similar disease occurs in adults older than 30 years. Usually, systemic EBV-positive LPD with hemophagocytosis in children occurs after the primary EBV infection, but in adult patients, the EBV might arise from a new infection or be activated from a latent infection.

Comment

It was generally agreed that the target for EBV infection in systemic EBV-positive T cell LPD of childhood in Asia is mainly CD8+ cytotoxic T cells [12]. However, EBV infections were also found in other lymphocyte subpopulations of CD4+ T cells and CD16+ NK cells although the frequency was much less than among CD8+ cells [17, 24, 25]. Therefore, the definition of “systemic EBV+ T cell LPD of childhood” should not be limited to T cells and should be expanded to include NK as well as T cells. The clonality of EBV-positive systemic T cell LPD of childhood is determined by various methods including assays for TCR gene rearrangements and cytogenetic abnormalities, and EBV genome terminal repeat investigation. The rate of cytogenetic abnormalities has been reported as 30 % [18]. Among patients with acute hemophagocytic lymphohistiocytosis, EB viral monoclonality was detected in more than 80 % of cases. Imashuku et al. analyzed EBV clonality by Southern blot analysis and EBV was monoclonal in 21 patients, biclonal in 2 patients, and non-clonal in 2 patients among 25 patients [18, 25]. TCR gene rearrangement has been detected in about 33–100 % of cases. By using Southern blot analysis, 17 among 25 patients showed monoclonality for TCR genes; ten patients for TCR beta chain gene, three patients for TCR-gamma chain gene, and two patients for both beta and gamma chain genes. Matsuda et al. reported that all six patients showed monoclonal peaks in TCRβ and/or TCRγ using European BIOMED-2 multiplex PCR. Ahn et al. showed monoclonal TCR-gamma gene rearrangement in 10 among 30 patients by conventional PCR method. [18, 2527].

The absence of clonality detected from TCR gene rearrangements does not necessarily mean a non-clonal disease but more likely indicates NK cell proliferation. Although EBV clonality is more reliable method to identify clonality of EBV-positive disease, EBV clonality is difficult to analyze routinely because the technique needs fresh samples with large amounts of DNA for Southern blotting. Importantly, clonality defined by TCR gene rearrangement and EBV gene terminal repeats has little clinical impact in terms of the outcomes for patients. The overall clinical, laboratory, and pathology findings are similar irrespective of T cell and EBV clonality [18].

Question 2: Is aggressive NK cell leukemia an NK cell counterpart of systemic T cell LPD of childhood?

In response to this question, the experts agreed that it might be true in young children but not in adult patients. The EBV-infected cells in cases of systemic T cell LPD in younger children are mainly CD8+ cytotoxic T cells, but in a minority of cases the EBV infects NK cells or CD4+ cells [25]. Systemic EBV+ NK cell proliferation in children can be called “aggressive NK cell leukemia.” Such EBV-positive T/NK cell LPD in childhood arises on a background of primary EBV infection.

Aggressive NK cell leukemia in adult patients shares clinical similarities with systemic T cell LPD in children, but there are some differences. Pathologically infiltrating cells in systemic T cell LPD of childhood are usually small and lack significant cytologic atypia, while aggressive NK cell leukemia in adult patients commonly show obvious cytologic atypia of tumor cells with significant cytogenetic abnormalities [4]. Aggressive NK cell leukemia in adult has no relation with primary EBV infection but caused by reactivation or reinfection of EBV in the elderly patients.

Chronic active EBV infection of T and NK cells

Question 3: What is the best nomenclature to describe the clinicopathological spectrum of chronic active EBV infections?

Regarding this question, some experts favored “EBV-positive T/NK cell LPD in childhood.” Others suggested the term “chronic active EBV disease” rather than “infection” because it is not a simple infection but an LPD induced by EBV infection. Finally, “chronic active EBV disease-type T/NK cell LPD” was adopted as the most appropriate term.

Comment

Chronic active EBV infection (CAEBV) of T and NK cell type is always associated with proliferation of T or NK cell to varying degrees [8, 10, 11]. The proliferating cells in cases of CAEBV frequently lack histological evidence of malignancy and can be polyclonal, oligoclonal, or monoclonal according to the stage of transformation [10]. Because CAEBV constitutes a continuous spectrum of EBV-infected T/NK cell proliferation, nomenclature to describe the pathological category of CAEBV is needed to facilitate communication among clinicians and pathologists. Ohshima et al. proposed a three-tier classification for CAEBV diseases [8]. Category A1 is polymorphic LPD with polyclonal proliferation of EBV-infected T or NK cells. Category A2 is polymorphic LPD with proliferation of monoclonal T or NK cells. Category A3 is monomorphic LPD of monoclonal T or NK cells and is equivalent to EBV-associated T/NK lymphoma/leukemia (Table 2). Experts agreed to accept this three-tier classification for the diagnosis of CAEBV disease. Categories A2 and A3 might correspond to systemic T cell LPD of childhood arising in patients with CAEBV disease in the current WHO classification [12].

Table 2 Pathological category of EBV-associated T/NK LPDs proposed by Ohshima et al. [8]

Hydroa vacciniforme and hydroa-like T cell lymphoma

Question 4: What are the diagnostic criteria that can distinguish HV from HV-like lymphoma?

The experts concluded that there are no reliable criteria to separate HV from HV-like T cell lymphoma because they constitute a continuous spectrum of diseases and proposed the nomenclature “Hydroa-like T cell LPD of childhood type.”

Comment

HV has been divided into two types based on the clinical features [28, 29]. The classic type is a self-limited form of disease characterized by the formation of vesicles on sun-exposed areas and follows a benign course, typically resolving in adolescence or young adulthood [29, 30]. Patients with the severe type tend to show more extensive skin lesions and these are often complicated by chronic active EBV infection. About half of patients with severe HV develop EBV-associated NK/T cell lymphomas in the skin or other organs 2–14 years after onset [29]. Hydroa-like T cell lymphomas develop in patients with long-standing hydroa vacciniforme [21, 29] but may also occur as a de novo disease [19, 31, 32]. Severe HV and HV-like T cell lymphoma show significant overlapping in their histopathology. As the disease progresses to overt malignancy, the cutaneous lesions are more extensive and deeper and the infiltrating cells show an increasing incidence of T cell monoclonality. Considering the continuous spectrum of HV-like T cell LPD, the distinction of HV-like T cell lymphoma from severe HV is arbitrary and very difficult. However, for clinical practice and investigations, the disease would be better to be stratified according to the severity, pathological changes, and clonality of cells. Consensus on the diagnostic criteria to define each category of HV-like T cell LPD should be made in the near future.

Conclusions

At the end of the meeting, the experts proposed a nomenclature for EBV-positive T/NK LPD of childhood as in Table 3. Since each category of EBV+ T/NK LPD in childhood has been defined based on the clinical characteristics, clinical features were considered as a part of the disease definition. Under the umbrella term “EBV-associated T/NK LPD,” which covers the entire spectrum of EBV-associated lesions in childhood from reactive to neoplastic process, the disease was divided into systemic and cutaneous forms. Systemic T/NK cell LPD of childhood is a fulminant disease associated with proliferation of polyclonal, oligoclonal, or monoclonal T or NK cells and aggressive NK cell leukemia in children can be included in this category. Chronic active EBV disease-type T/NK cell LPD is a systemic illness lasting more than 3 months with histopathological evidence of major organ involvement and increased levels of EBV RNA or proteins in affected tissues or blood. CAEBV disease-type T/NK cell LPD is divided into three groups based on the histology and clonality of T or NK cells. A monoclonal EBV+ T/NK cell proliferation with the clinical features of CAEBV rather than the fulminant course of systemic EBV-positive T/NK cell LPD of childhood is best regarded as CAEBV disease-type T/NK cell LPD.

Table 3 Proposed nomenclature of EBV-positive T/NK LPD of childhood from consensus meeting of the 4th Asian Hematopathology Workshop

Mosquito bite hypersensitivity and hydroa vacciniforme-like T cell LPD constitute cutaneous form of EBV-positive LPD in childhood. Since two entities are distinct in the clinical characteristics and the lineage of infiltrating cells, they are worth specifically named in the classification of EBV-positive T/NK LPD.