Search tips
Search criteria 


Logo of bloodcancerLink to Publisher's site
Blood Cancer J. 2016 May; 6(5): e417.
Published online 2016 May 6. doi:  10.1038/bcj.2016.26
PMCID: PMC4916295

The microenvironment of classical Hodgkin lymphoma: heterogeneity by Epstein–Barr virus presence and location within the tumor

Classical Hodgkin lymphoma (cHL) is characterized by a low number of malignant cells, called Hodgkin Reed–Sternberg (HRS) cells, surrounded by an abundant infiltrate of immune cells. Presence of T helper (Th)2 and regulatory T cells (Treg) in the infiltrate was shown by staining of T-cell subtype specific transcription factors in cHL tissue samples1 and by expression of Th2- and Treg-specific cytokines in sorted T-cell subsets.2 However, one study revealed a predominant Th1 subpopulation in the microenvironment of cHL based on expression of CXCR3 by flow cytometry.3 Differences between these studies may potentially be explained by heterogeneity between Epstein–Barr virus (EBV)-positive and -negative cHL cases and by differences in cell composition in relation to proximity of tumor cells.

In this study, we compared the composition of the reactive infiltrate of EBV+ (n=7) and EBV− cHL (n=7) with each other and with reactive lymph nodes (RLNs) for 46 specific cell types (Supplementary Table S1). In addition, we made a distinction between cells that are located in the close vicinity of tumor cells and cells that are further away. The characteristics of the study cohort has been described in Table 1. The gender distribution and age range of the HL patients are comparable with the RLN cohort. There is no significant difference in ages between the EBV+ and EBV− patients. The majority of the HL patients have nodular sclerosis (NS) histology, within the EBV+ cHL group relatively more mixed cellularity (MC) cases are found.

Table 1
Characteristics of the study cohorts

We did not find any significant differences between HL and RLN in the main cell populations (B cells, T cells, natural killer (NK) cells and macrophages)(Supplementary Table S2). In contrast to previous studies using immunohistochemistry (range 3–57%),4 we observed relatively low percentages of macrophages. This might be caused by a less optimal isolation of macrophages from fresh tissues, resulting in an underrepresentation of these cells in our flow analysis.

Analysis of different T-cell subpopulations revealed significantly more GITR+CD25+ and FoxP3+CD25+ Treg cells in the CD4+ T-cell subset in both EBV+ and EBV− cHL compared with RLNs (Figures 1a and b). This increase in Treg cells is consistent with a previously published study.5 We observed no significant difference for EBV+ and EBV− cHL in the percentage of Tregs, suggesting that suppression of the immune response is important in both EBV+ and EBV− cHL. FoxP3+CD25+CD4+ Tregs were upregulated in EBV+ HL cases by immunohistochemistry.6 Moreover, upregulation of messenger RNA (mRNA) levels for Treg-associated markers and cytokines has also been described in EBV+ cHL compared with RLN and EBV− cHL, especially for T regulatory 1 cells, that produce interleukin-10.7 Together these findings indicate that Tregs are important in all cHL cases, albeit more pronounced in the microenvironment of EBV+ HL.

Figure 1
Significantly different immune cell populations in cHL. Optimized concentrations of fluorochrome-antibody conjugates were applied on cell suspensions of cHL tissue and analyzed by flow cytometry. (a–f) Significant differences in immune cell populations ...

The percentage of CD56+CD16+ NK cells was significantly lower in EBV− cHL compared with RLN and EBV+ cHL (Figure 1c). In addition, we observed significantly more CD69+, granzyme B+ and TIA-1+ cells in the CD8+ T-cell subset in EBV+ cHL compared with RLN and EBV− cHL (Figures 1d–f). Thus, EBV status has a positive correlation with the percentage of NK cells and activated CD8+ T cells. This suggests that the presentation of EBV-derived antigenic peptides within human leucocyte antigen class I molecules triggers antiviral immune responses. The number of NK cells can increase upon recognition of virus by toll-like receptors. The higher numbers of cytotoxic CD8+ T cells in EBV+ cHL indicates a Th1 response as Th1 cells stimulate a cellular-immune response and activate CD8+ T cells. In contrast to the findings of Greaves et al.,3 we did not find significant increases in the percentage of CXCR3+ Th1 cells within the CD4+ cells in either EBV+ or in EBV− cHL.

The combined presence of Tregs and activated CD8+ T cells raises the questions on how they can coexist in cHL and where they are located in respect to the tumor cells. To answer these questions, we used CD26 to discriminate between T cells in the area of the HRS cells, that is, the CD26− T cells, and the T cells that are not in close contact with the HRS cells, that is, the CD26+ T cells.2 To confirm the relevance of using CD26 as a marker, we first stained CD26 in the cHL tissue sections. In the NS and lymphocyte-rich subtypes of cHL, CD26− cells were indeed located in the close vicinity of tumor cells, whereas CD26+ cells were not observed in the tumor cell areas. In MC, the CD26+ cells were much more dispersed and were also found close to the tumor cells. Based on these staining results, we excluded the MC cases from the subsequent analyses.

We found significantly more CD69+ Th cells in the CD26− T-cell subset indicating that these cells are more abundant in the tumor cell area (Figure 1g). This suggests triggering of an immune reaction in these cells because CD69 is an early activation marker. A similar pattern was observed in EBV+ as well as EBV− cHL. Furthermore, we also observed more FoxP3+-suppressive Th cells in the tumor cell area in both EBV+ and EBV− cHL (Figure 1h). These findings are consistent with increased mRNA levels of Treg-associated genes (CTLA4, IL2RA, TNFRSF4 and CCR4) in CD4+CD26− T cells in cHL compared with RLN.2 So, the Tregs reside in the direct vicinity of the HRS cells and might inhibit an effective antitumor immune response. There were less CD25+ CD8+ cells around the tumor cells than outside the tumor cell area (Figure 1i). So, although activated CD8+ cells are present in the microenvironment of cHL, these cells are not present in the tumor-cell-rich areas and thus not effective.

In conclusion, we showed that the main difference in the composition of the microenvironment between EBV+ and EBV− cHL patients is caused by increased numbers of both CD8+ T cells and NK cells. High numbers of CD4+ Tregs directly surrounding HRS cells imply that they have an important role in the failure to induce an effective immune response against the HRS cells of cHL.


The authors declare no conflict of interest.


Supplementary Information accompanies this paper on Blood Cancer Journal website (

Supplementary Material

Supplementary Information


  • Schreck S, Friebel D, Buettner M, Distel L, Grabenbauer G, Young LS et al. Prognostic impact of tumour-infiltrating Th2 and regulatory T cells in classical Hodgkin lymphoma. Hematol Oncol 2009; 27: 31–39. [PubMed]
  • Ma Y, Visser L, Blokzijl T, Harms G, Atayar C, Poppema S et al. The CD4+CD26- T-cell population in classical Hodgkin's lymphoma displays a distinctive regulatory T-cell profile. Lab Invest 2008; 88: 482–490. [PubMed]
  • Greaves P, Clear A, Owen A, Iqbal S, Lee A, Matthews J et al. Defining characteristics of classical Hodgkin lymphoma microenvironment T-helper cells. Blood 2013; 122: 2856–2863. [PubMed]
  • Tan KL, Scott DW, Hong F, Kahl BS, Fisher RI, Bartlett NL et al. Tumor-associated macrophages predict inferior outcomes in classic Hodgkin lymphoma: a correlative study from the E2496 Intergroup trial. Blood 2012; 120: 3280–3287. [PubMed]
  • Bosler DS, Douglas-Nikitin VK, Harris VN, Smith MD. Detection of T-regulatory cells has a potential role in the diagnosis of classical Hodgkin lymphoma. Cytometry B Clin Cytom 2008; 74: 227–235. [PubMed]
  • Assis MC, Campos AH, Oliveira JS, Soares FA, Silva JM, Silva PB et al. Increased expression of CD4+CD25 +FOXP3+ regulatory T cells correlates with Epstein-Barr virus and has no impact on survival in patients with classical Hodgkin lymphoma in Brazil. Med Oncol 2012; 29: 3614–3619. [PubMed]
  • Morales O, Mrizak D, François V, Mustapha R, Miroux C, Depil S et al. Epstein-Barr virus infection induces an increase of T regulatory type 1 cells in Hodgkin lymphoma patients. Br J Haematol 2014; 166: 875–890. [PubMed]

Articles from Blood Cancer Journal are provided here courtesy of Nature Publishing Group