Introduction
Epstein Barr virus (EBV) is a human herpesvirus acquired in early life that establishes itself as a latent infection in memory B cells for life [
1]. Its ability to transform the B cells has been associated with various specific lymphoma subtypes such as Burkitt, classic Hodgkin (cHL), plasmablastic, and diffuse large B cell lymphoma (DLBCL) [
2]. Recently, in several EBV-associated B cell lymphomas, traces of EBV infection in cases originally considered as negative by EBERs in situ hybridization (ISH), the gold standard to detect EBV in lymphomas, were described. The presence of those traces provided the basis to explain the hit-and-run theory. According to this hypothesis, the viral infection in the B cell initially triggers the cellular transformation, and then, once the malignant process has been established, the lymphoma clone loses the EBV genome over time. Given the fact that the EBV traces may remain, some studies suggest that the virus could be involved in the pathogenesis of these lymphomas [
3‐
5]. Recently, our group also proved the presence of traces of EBV in patients with DLBCL from Argentina, but suggested, since higher expression of the most important oncogenic protein, as well as increased viral load, was observed in cases with EBERs + cells by conventional ISH, that traces of EBV could not display a key role in DLBCL pathogenesis [
6].
EBV+ DLBCL not otherwise specified (NOS) is a new entity confirmed by the World Health Organization (WHO) in 2017, with a high incidence in elderly patients, and at least 80% of the tumor cells positive for EBERs ISH [
7]. In this new entity, it has been described that the virus may contribute to some alterations in the tumor microenvironment, such as the dysregulation of the programmed death ligand 1 (PDL1) [
8].
The tumor microenvironment (TME) plays a key role in lymphomagenesis and is composed of T lymphocytes, macrophages, and natural killer (NK) cells, as well as stromal cells, blood vessels, and extracellular matrix (ECM) [
9]. The expression of the molecules that induce tolerance in the TME of EBV+ DLBCL leads to an immune-tolerogenic microenvironment [
10,
11]. The expression of PDL1, T cell immunoglobulin and mucin domain-containing protein-3 (TIM3), and lymphocyte activation gene-3 (LAG3) were described in DLBCL, as immune escape mechanisms to promote lymphomagenesis [
12]. TIM3 and LAG3 are expressed in the surface of CD4 + and CD8 + T cells, to induce the impairment of the cell function, and, in consequence, to ultimately promote immune evasion. On the other hand, PDL1 can be expressed in tumor cells, as well as in the microenvironment, reducing immune cell function upon ligation to its ligand, PD1, expressed in T cells. Furthermore, it was described that PDL1 expression on the surface of tumor cells could be strongly associated with poor prognosis in patients with DLBCL [
13,
14].
On the other hand, there are a few studies that explore tumor-associated macrophages (TMAs) in DLBCL. Macrophages are divided into at least M1 and M2 subtypes, depending on the expression of CD68 and CD163 markers, respectively, with opposing effector functions [
15]. Riihijärvi. S et al. demonstrated a positive correlation between the protein and gene expression of macrophages markers in patients with DLBCL [
16]. Furthermore, a high levels of CD68 protein and its RNA detected by gene expression approach in patients treated with chemoimmunotherapy were associated with favorable progression free survival (PFS) and overall survival (OS) in DLBCL [
16,
17]. In contrast, a recent study described high levels of CD163 + M2 macrophages in EBV+ DLBCL, also associated with inferior outcomes [
10].
Gene expression profiling (GEP) studies are used to quantitatively analyze factors related to the tumor and TME with prognosis and/or biological importance, and, specifically in DLBCL, it was used to classify distinct DLBCL molecular subtypes based on the cell of origin (COO) [
18]. Furthermore, a customized platform that includes genes associated with immune response expressed by the different microenvironmental and neoplastic components was used to identify markers in peripheral T cell lymphoma (PTCL) and to recognize the expression of B cell genes for angioimmunoblastic T cell lymphoma (AITL). This approach was assessed in routinely processed paraffin-embedded samples, providing a more specific diagnosis and prognosis [
19].
As described above, the role of TME is still under discussion in this new entity, EBV+ DLBCL, NOS, and it is still unexplored if the presence of EBV could trigger alterations at the TME, inducing a first “hit”, followed by the “run” after triggering, for example, a tolerogenic environment, as proposed by the hit-and-run theory. Under this hypothesis, once the alteration has been established, traces of EBV could be detected [
5,
6]. In addition, in Argentina EBV-associated lymphomas have a higher incidence in children < 10 years, and almost 90% of patients are seropositive by the age of 3 years [
20]. Furthermore, since our group previously suggested that traces of EBV could not display a key role in DLBCL pathogenesis, the aim of this study was to analyze the expression of immune response genes in the TME to disclose the role of the virus and its traces in DLBCL.
Discussion
Recently, it has been suggested that EBV may be involved in the pathogenesis of lymphomas in more cases than originally considered, based on the detection of viral traces sensitive techniques [
3‐
5]. Furthermore, this hypothesis was also proposed for gastric cancer, where the methylation changes were studied as a mechanistic framework for how EBV can act in a hit-and-run manner [
26]. However, even though EBV was detected by sensitive methods in patients with DLBCL from Argentina in more cases than originally considered, a higher expression of the most important oncogenic protein, LMP1, as well as increased viral load, was observed in cases with EBERs+ cells by conventional ISH, suggesting that traces of EBV could not display a key role in DLBCL pathogenesis [
6]. Given the fact that EBV-associated lymphomas have a higher incidence in children < 10 years in Argentina, and almost 90% of patients are seropositive by the age of 3 years [
20], our aim was to evaluate the local immune response in this population, to enlighten the role of traces of EBV infection in the pathogenesis of DLBCL.
Through the gene expression assay for immune response genes, no significant differences neither between the EBV+ groups compared to the EBV− cases nor between the cases that expressed viral traces compared to the cases without viral traces were demonstrated, in spite of the low number of patients included, one of the limits of this study. However, the slight increase in the expression of the
CD8A gene prompted us to perform the analysis for cell types. Remarkably, the EBV+ DLBCL cases displayed an enhanced gene expression for cell types such as CD8 T and cytotoxic cells. This finding was not unexpected, since our group in a previous study showed an increase in CD8 + T cells and granzyme B + cytotoxic effector cells in EBV+ DLBCL NOS, associated with a tolerogenic microenvironment [
27], that is the reason why in this study this increase was not confirmed by IHC. This increase in the expression of markers for CD8 T and cytotoxic cells for EBV+ cases but not for the cases that expressed viral traces might indicate the presence of traces could not be able to trigger this specific scenario. Alternatively, it could be assumed that an initial “hit” followed by the “run” is not sufficient to modify the local immune response, and the presence of high numbers of EBV-infected cells is required to induce cytotoxic response.
It is well known that the EBV+ DLBCL have a tolerogenic environment, characterized by the expression of PD1, TIM3, and LAG3 in tumor-infiltrating lymphocytes [
10‐
14], and PDL1 in different types of cells, such as macrophages [
14,
28,
29]. In contrast, the lack of significant differences in the expression of LAG3, TIM3, and PDL1 at the microenvironment may reveal that, specifically in this series, neither the EBV nor its traces could have influenced in the expression of these particular tolerogenic proteins at the tumor microenvironment.
The slight increase observed in
PDL1 gene expression
(CD274) analyzed by Nanostring reflects its expression at the microenvironment, but also in tumor cells, another limit of this specific approach. In fact, increased PDL1 expression in tumor cells was related to EBV− presence in DLBCL [
28,
30]. The expression of PDL1 in tumor cells could be consequence of PDL1 genetic alterations, which were demonstrated in EBV+ DLBCL [
31], but also as a result of the activation by LMP1 and EBNA2 viral latent oncogenic proteins [
32,
33]. In line with this, in the analyzed cohort, the expression of PDL1 only in tumor cells turned out to be significantly higher only for EBV+ cases, indicating that the virus, and its oncogenic proteins, could be involved in this increased expression. However, the traces of EBV infection might not be responsible for PDL1 upregulation.
The presence of different factors can induce macrophage polarization toward M1 or M2 profiles. M1, characterized by higher CD68 expression, is induced by a proinflammatory phenotype with tumoricidal activity. Instead, the M2 profile, with higher CD163 expression, promotes tissue repair and Th2 immune response and favors a tumorigenic propitious environment, promoting neoangiogenesis, tissue invasion, and metastasis [
34]. In the context of EBV-associated lymphomas, M2 polarization was described in Burkitt lymphoma [
35]. In contrast, M1 polarization was described in EBV-associated pediatric Hodgkin lymphoma [
36,
37]. Concerning DLBCL, both EBV+ and EBV− cases exhibit a prevalence of M2 polarized macrophages so far [
38,
39]. In contrast, in the whole series the intermediate polarization pattern confirmed by CD68/CD163 ratio < 1.5 predominated. Moreover, the small increase in CD68 gene expression observed in EBV+ DLBCL cases in our cohort was confirmed by immunohistochemical CD68 expression. Nevertheless, this increase was lost when only cases with traces of EBV infection were evaluated, indicating that they have no influence in CD68 upregulation.
In summary, this study provides further evidence of the role of traces of EBV infection in the pathogenesis of DLBCL, in a cohort from a population with a high incidence of EBV infection in children. Even though its detection may support the hit-and-run hypothesis, it seems like the immune response markers analyzed in this series, such as CD8, cytotoxic T cells, PDL1 and CD68, only are increased when EBERs is expressed in more than 20% of tumor cells, defined as EBV+ DLBCL. When only traces are detected by sensitive methods, they might not have influence in immune response markers, indicating that perhaps the initial “hit” is not enough to sustain the changes in the local milieu. Be as it may, further studies are needed to shed light on EBV involvement in the lymphomagenesis process in DLBCL.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.