Tel: +49 30 5681 3750; Fax: +49 30 5681 3753
Tumor microenvironment composition in pediatric classical Hodgkin lymphoma is modulated by age and Epstein-Barr virus infection
Article first published online: 21 DEC 2011
Copyright © 2011 UICC
International Journal of Cancer
Volume 131, Issue 5, pages 1142–1152, 1 September 2012
How to Cite
Barros, M. H. M., Vera-Lozada, G., Soares, F. A., Niedobitek, G. and Hassan, R. (2012), Tumor microenvironment composition in pediatric classical Hodgkin lymphoma is modulated by age and Epstein-Barr virus infection. Int. J. Cancer, 131: 1142–1152. doi: 10.1002/ijc.27314
- Issue published online: 27 JUN 2012
- Article first published online: 21 DEC 2011
- Accepted manuscript online: 25 OCT 2011 12:18AM EST
- Manuscript Accepted: 26 SEP 2011
- Manuscript Received: 26 AUG 2011
- INCT para Controle do Câncer. Grant Number: CNPq 573806/2008-0
- FAPERJ. Grant Number: E26/170.026/2008 (Brazil)
- Swissbridge Foundation (Switzerland)
- Grant sponsor:Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ). Grant Number: E26/111.432/2010
- CAPES-DAAD program (Brazil-Germany)
- Hodgkin lymphoma;
- Epstein-Barr virus;
- tumor microenvironment;
Classical Hodgkin lymphoma (cHL) is characterized by a small number of neoplastic cells in a background of reactive cells. Children and adults differ in constitution and functionality of the immune system and it is possible that there may be age-related differences in tumor microenvironment composition in cHL. One hundred children with pediatric cHL were studied. Tumor-infiltrating lymphocytes were analyzed by immunohistochemistry (IHC) and image analysis. Epstein-Barr virus (EBV) status was determined by EBER-specific in situ hybridization and IHC. Results were analyzed in the context of age-group, histological characteristics and clinical follow-up. EBV-status was not associated with age-group. Children <10 years and EBV+ cases were characterized by a more intense T cell infiltrate, exhibiting a cytotoxic/Th1 profile, characterized by higher numbers of CD3+, CD8+, TIA1+ and TBET+ lymphocytes. Extranodal disease (p = 0.016) and high number of GranzymeB+ lymphocytes (p = 0.04) were independently associated with reduced progression-free survival (PFS). Yet, in EBV+ cases, improved outcome was observed in cases with low numbers of FOXP3+ lymphocytes (p = 0.046), FOXP3/CD8 ratio < 1 (p = 0.021) and TBET/CMAF ratio < 1 (p = 0.017). By contrast, in EBV− cases, poor survival was observed in cases with extranodal disease (p = 0.028), MC subtype (p = 0.009) and high numbers of TIA1+ (p = 0.044) and GranzymeB+ (p = 0.04) lymphocytes. The results suggest that in EBV+ cHL an effective immune response directed against viral or tumor antigens may be triggered in the tumor microenvironment and that physiological and age-related changes of the immune system may also modulate the tumor microenvironment in pediatric cHL.
Classical Hodgkin lymphoma (cHL) is characterized by a small number of neoplastic, Hodgkin and Reed-Sternberg (HRS) cells, in a background of inflammatory non-neoplastic cells, mainly B and T cells.1 The clinical and pathological characteristics of cHL are thought to result from an immune dysfunction, characterized by the deregulated expression of several cytokines and chemokines,2 and an inadequate immune response due to poor immunogenicity of HRS cells, the immunosuppressive effects exerted by the tumor cells and/or the poor response of the host immune system.2
At a local level, the tumor microenvironment is maintained by a crosstalk between malignant and reactive cells in the microenvironment, mediated by cytokines and chemokines expressed by both, HRS and inflammatory cells.2 Increasing evidence of the importance of tumor microenvironment in the pathogenesis of and clinical response to cHL has come from immunohistochemical and molecular studies.2–4 Particularly, the number of several intratumoral cell populations, such as cytotoxic and regulatory (Treg) T cells,2, 5, 6 or the ratio between these populations7–9 have been described as being associated with therapeutic response.
Up to now, however, all these studies have been conducted in adult patients or a mixture of adolescents and adults. It is possible that children and adults differ in their response to the tumor, not only with respect to the recruited cell types, but also to the functional status of these cells. In healthy individuals, the numbers of T and B lymphocytes in the peripheral blood differ between children and adults.10–13 Furthermore, in adults aging of the immune system characterized by the accumulation of CD8+CD28− cells may contribute to the shaping of the immune response to tumor or viral antigens.14–16
The presence of Epstein-Barr virus (EBV), which clonally infects HRS cells in a proportion of cases, is another variable capable of modulating the tumor microenvironment by inducing the expression of cytokines involved in cell migration and of viral immunogenic proteins.17–19 In fact, molecular profiling studies have identified a distinct immune molecular signature for EBV-associated cHL.18, 20
In this study, we have analyzed the composition of the tumor microenvironment and its prognostic impact in a large series of pediatric cHL comprising comparable numbers of EBV+ and EBV− cases.
Material and Methods
One hundred children and adolescents (up to 18 year old) diagnosed with cHL at the Instituto Nacional do Câncer (INCA, Brazil) between 1999 and 2006 were included in this study, based on availability of formalin-fixed paraffin-embedded tissue and complete clinical data for retrospective analysis. All patients were negative for the human immunodeficiency virus (HIV) infection. Patients were classified in two age groups (≤10 years vs. >10 years) to investigate microenvironment characteristics in younger vs. older children.10–13, 21 Pretreatment factors evaluated were disease stage according to the Ann Arbor staging criteria,22 presence of B symptoms and mediastinal mass, number of involved anatomic areas (IAA) (nodal and extranodal), presence of anemia (children <12 years: Hb <12 g/dl; male 12–18 years: Hb <13.5 g/dl; female 12–18 years: 11.5 g/dl), leukopenia (<4,000 leukocytes /mm3) and lymphopenia (<800 lymphocytes /mm3). Evaluation of IAA was performed as previously described.20 Children with Stages I, IIA or IIIA were classified as favorable clinical presentation and children with Stages IIB, IIIB or IV were classified as unfavorable clinical presentation, according pediatric oncology guidelines.23 All patients were treated according pediatric protocols and received adriblastine-based chemotherapy:24, 25 53 children (53%) were treated according to the ABVD-protocol (adriamycin, bleomycin, vinblastine and dacarbazine) and 47 children (47%) according to the German pediatric protocol HD90, consisting of OPPA (vincristine, procarbazine, prednisone and doxorrubicine) for girls, OEPA (vincristine, procarbazine, prednisone and doxorrubicine) for boys and COPP (cyclophosphamide, vincristine, procarbazine and prednisone) for both when they were not presenting with early stage disease (Stages I and IIA).25 Radiotherapy was administered only when was proposed by the protocols.24, 25 This study was approved by the INCA Ethics Committee.
Diagnosis of cHL was established by morphologic criteria according to the WHO classification.1 All cases were independently reviewed by two pathologists (MHMB and GN); discordant cases were discussed at a double-headed microscope. In cases no agreement was reached, cases were designated “unclassifiable”. The interfollicular pattern was defined as the presence of frequent residual follicles or extensive follicular hyperplasia with interfollicular tumor infiltrates.26 All histological parameters were evaluated in conventional H&E-stained slides.
Tissue microarray design
A Tissue arrayer device (Beecher Instrument, SilverSpring, MD) was used to assemble the tissue microarray (TMA) blocks. From each case, two 1-mm-diameter cores, selected from two different representative tumor areas, were included.
TMA blocks were sectioned at a thickness of 3 μm. Sections were dried for 12 hr at 56°C. Dewaxed paraffin sections were rehydrated and submitted to endogenous peroxidase blockade. Antigen retrieval was performed by heat treatment in a pressure-cooker for 1 min. Buffers used for antigen retrieval and primary antibodies are listed in the Supporting Information Table S1. Briefly, the antibodies used were anti- CD3, CD4, FOXP3, CD8, TIA1, Granzyme B and CD20. Anti-TBET and anti-CMAF antibodies were used for characterization of Th1 and Th2 cells, respectively, as previously reported.27 Proliferative activity of reactive non-neoplastic lymphocytes was evaluated by Ki67 immunostaining.
After incubation with primary antibody, immunodetection was performed using labeled streptavidin biotin (LSAB) visualization system (Dako, Glostrup, Denmark), Envision kit (Dako) or ZytoChem Plus HRP polymer kit (Zytomed Systems, Berlin, Germany) (Supporting Information Table S1), employing diaminobenzidine (DAB) chromogen as substrate. Sections were counterstained with hematoxylin. External and internal controls included in the TMA were taken into consideration to interpret stainings.
Latent EBV infection was determined in all patients by in situ hybridization (ISH) with fluorescein-conjugated probes for EBV-encoded RNAs (EBER-ISH) and by IHC against latent membrane protein 1 (LMP1) as described.20
Computer-assisted microscopical analysis
For lymphocyte subset evaluation, each core was photographed using AxioCam MRc camera (Zeiss, Germany) at a 200× magnification. The numbers of labeled lymphocytes were determined per 1 mm2 using the image analysis software HISTO (Biomas, Elangen, Germany). Germinal centers without colonization by neoplastic cells, when present in the core, were not considered in the cell count.
For each lymphocyte subset, the 25th and 50th percentiles were used to categorize the intensity of the lymphocyte infiltration. Cases displaying no labeled lymphocytes in both cores were considered not evaluable for technical reasons.
Pearson's chi-square and Fisher's exact test were used to test association between dichotomous variables. Mann-Whitney test was used to test association between dichotomous and continuous variables, while Spearman's correlation was used to test association between continuous variables. Differences were considered significant at p < 0.05 in two-tailed tests. Primary treatment was defined as a failure if the lymphoma had progressed at any time after the initiation of therapy; treatment success was defined as the absence of progression or relapse. Progression-free survival (PFS) was the interval (in months) from diagnosis to progression at any time, relapse from complete response, or initiation of new, previously unplanned treatment or to the last follow-up in the patients with treatment success. Overall survival (OS) refers to the interval (in months) from the diagnosis to death or last follow-up. Survival distributions were estimated by the Kaplan-Meier method and differences were compared using log-rank test. Multivariate Cox proportional hazard regression method was used to determine the independent prognostic value of statistically significant variables in univariate analyses. Data were analyzed using Statistical Package for the Social Sciences 13.0 (SPSS).
Clinico-pathological parameters and EBV status
The age at diagnosis ranged from 3 to 18 years (median 14 years), with a male:female ratio of 1.7:1 (64 males and 36 females). Similar numbers of patients were allocated to favorable and unfavorable clinical presentation groups (52.6% vs. 47.4%).The number of IAA ranged from 1 to 9 (median 3).
NS was the predominant subtype (69/100, 69%), followed by mixed cellularity (MC) (23/100, 23%). The presence of granulomas was observed in 37% of cases (37/100 cases), predominantly in MC subtype (14/23 MC cases with granulomas vs. 21/69 NS cases with granulomas, p = 0.007, χ2). Granulomas were found preferentially in the young age-group (15/27 children ≤ 10 years had granulomas, and 22/73 children > 10 years had granulomas. p = 0.019, χ2). There was no association of this histological feature with EBV status or other clinical-histological variables.
Overall, EBV was detected in 44.8% cases, with the MC subtype (15/22, 68.1%) showing a higher association than the NS subtype (28/74, 37.8%) (p = 0.012); there was no association with age-groups (≤ vs. >10 years; p = 0.185). The main clinical and histological characteristics are listed in Table 1.
Lymphocyte subsets in pediatric cHL
CD3+ T cells as well as T-cell subsets showed no specific distribution in the tumor microenvironment, being sometimes observed in the vicinity of HRS cells in a rosette-like fashion, but without predominance of this feature (Fig. 1). The CD20+ B cells were distributed uniformly in the tumor microenvironment, without specific localization.
A detailed summary of the numbers of each lymphocyte population per mm2 is given in the Table 2. A higher number of CD3+ T cells in comparison to CD20+ B lymphocytes was observed (median 645 cells/mm2 vs. 196 cells/mm2, respectively). However, CD20+ B cells were observed at least as frequently as any T cell subset individually [CD4+ (median 155 cells/mm2), FOXP3+ (median 49 cells/mm2), CMAF+ (median 68 cells/mm2), TBET+ (median 32 cells/mm2), CD8+ (median 143 cells/mm2), TIA1+ (median 69 cells/mm2) and Granzyme B+ (median 11 cells/mm2)] (Fig. 2).
Ki67 immunostaining was evaluated in the small mature benign lymphocytes. A large variation was observed regarding the number of non-neoplastic Ki67+ lymphocytes (3–352 cells/mm2, median: 91 cells/mm2). These lymphocytes showed no specific distribution in the tumor microenvironment. No association with histological features or with EBV-status was found. A direct correlation (Spearman's correlation) was observed between the number of benign Ki67+ lymphocytes and the number of FOXP3+ (p = 0.042), CD8+ (p < 0.0005), TIA1+ (p = 0.015), Granzyme B+ (p < 0.0005), TBET+ (p < 0.0005) and CMAF+ (p < 0.0005) lymphocytes (Supporting Information Fig. S1).
Lymphocyte subsets are associated with clinical and histological features
The younger age group (≤10 years) was characterized by a more intense T cell infiltrate, exhibiting a cytotoxic/Th1 profile, characterized by higher numbers of CD3+ (p = 0.025; Mann-Whitney test), CD8+ (p = 0.016, Mann-Whitney test), TIA1+ (p = 0.008, Mann-Whitney test) and TBET+ (p = 0.045, Mann-Whitney test) lymphocytes, whereas in older children (>10 years) FOXP3+, CMAF+ and CD4+ cells were more prevalent (Table 3).
In general, favorable clinical parameters were associated with a more intense T lymphocyte infiltrate. Stages I/II were associated with high numbers of CD3+ (median 748 cells/mm2 vs. 510 cells/mm2 in III/IV stages, p = 0.004, Mann-Whitney test), CD4+ (median 209 cells/mm2 vs. 99 cells/mm2 in III/IV stages, P = 0.005, Mann-Whitney test) and CMAF+ (median 94 cells/mm2 vs. 29 cells/mm2 in III/IV stages, p = 0.003, Mann-Whitney test) lymphocytes. Conversely, children with ≥4IAA exhibited low numbers of CD3+ (median 457 cells/mm2 vs. 752 cells/mm2 with <4IAA, p < 0.0005, Mann-Whitney test), CD4+ (median 91 cells/mm2 vs. 209 cells/mm2 with <4IAA, p = 0.027, Mann-Whitney test), CMAF+ (median 44 cells/mm2 vs. 94 cells/mm2 with <4IAA, p = 0.032, Mann-Whitney test) and CD8+ (median 111 cells/mm2 vs. 183 cells/mm2 with <4IAA, p = 0.008, Mann-Whitney test) lymphocytes in the tumor microenvironment.
MC subtype was associated with high numbers of CD8+ (p = 0.049, Mann-Whitney test), TIA1+ (p = 0.036, Mann-Whitney test) and Granzyme B+ (p = 0.019, Mann-Whitney test) lymphocytes, typical of a cytotoxic profile, when compared with all other types. This was independent of EBV status. Moreover, as previously described,28 MC was characterized by low numbers of CD4+ lymphocytes (p = 0.048, Mann-Whitney test), high numbers of CD20+ lymphocytes (p = 0.046, Mann-Whitney test) and CD4/CD20 ratio <1 (p = 0.005, Fisher's test).
Conversely, NS subtype was characterized by high numbers of CD4+ (p = 0.037, Mann-Whitney test) and low numbers of cytotoxic CD8+ (p = 0.041, Mann-Whitney test), TIA1+ (p = 0.032, Mann-Whitney test) and Granzyme B+ (p = 0.016, Mann-Whitney test) lymphocytes when compared with MC subtype (Supporting Information Table S2).
The interfollicular pattern was associated with high number of CD8+ (median 272 cells/mm2 vs. 143 cells/mm2 in absence of this pattern, p = 0.024, Mann-Whitney test), TIA1+ (median 152 cells/mm2 vs. 62 cells/mm2 in absence of this pattern, p = 0.037, Mann-Whitney test) and CD20+ (median 360 cells/mm2 vs. 168 cells/mm2 in absence of this pattern, p = 0.003, Mann-Whitney test) lymphocytes.
EBV+ cases displayed a predominantly cytotoxic microenvironment profile characterized by high numbers of CD8+ (p = 0.001, Mann-Whitney test), TIA1+ (p = 0.002, Mann-Whitney test), Granzyme B+ (p = 0.002, Mann-Whitney test), as well as TBET+ (p = 0.031, Mann-Whitney test) lymphocytes. There was no difference in relation to the number of Treg cells (p = 0.1, Mann-Whitney test). Detailed results are shown in Table 3.
Characteristics of the tumor microenvironment are associated with overall survival and prognosis
OS at 60 months was 89.4%. Clinical features associated with poor OS were absence of EBV-association (p = 0.029, log-rank) and low CD3+ lymphocytes number (<25th percentile) (p = 0.04, log-rank). A trend for unfavorable survival was observed in cases with low CD20+ lymphocytes number (<25th percentile) (p = 0.06, log-rank) (Supporting Information Fig. S2). All variables lost significance in the multivariate Cox analysis.
PFS at 60 months was 78.6%. Clinical features associated with poor PFS were extranodal disease (p = 0.028, log-rank) and leukopenia (p = 0.034, log-rank) (Table 1). EBV-associated cases showed a trend to a better PFS (p = 0.082, log-rank). Regarding the tumor microenvironment variables, worse PFS was observed in children with granulomas (p = 0.026, log-rank) (Table 2), low CD3+ lymphocytes number (<25th percentile) (p = 0.037, log-rank) and high Granzyme B+ lymphocytes number (>25th percentile) (p = 0.045, log-rank) (Table 2, Fig. 3).
A trend towards a worse PFS was observed for MC subtype (p = 0.057, log-rank), high TIA1+ lymphocytes number (>25th percentile) (p = 0.08, log-rank) and high number of Ki67+ non-neoplastic lymphocytes (>25th percentile) (p = 0.093, log-rank) (Table 2) (Fig. 3). In the Cox regression, extranodal disease (p = 0.016) and the number of Granzyme B+ lymphocytes (p = 0.04) maintained an independent prognostic impact (Supporting Information Table S3).
EBV infection of HRS cells may modulate the prognostic role of the tumor microenvironment
Given that EBV may be able to modulate the tumor microenvironment composition and local EBV-specific immunity,18, 29 the impact of these variables on survival was investigated according to EBV-status. In the EBV-negative group, a worse PFS was observed for MC subtype (p = 0.009, log-rank), and for cases with TIA1+ lymphocytes > 25th percentile (p = 0.044, log-rank) and Granzyme B+ lymphocytes > 25th percentile (p = 0.04, log-rank). Although extranodal disease is not a microenvironment variable, it is important to note that the observed worse PFS for patients presenting with extranodal disease was restricted to EBV-negative patients (p = 0.028, log-rank) (Supporting Information Table S4). In Cox regression, MC subtype (p = 0.005), number of cytotoxic cells (p = 0.024) and extranodal disease (p = 0.002) maintained their prognostic impact in EBV-negative cases (Supporting Information Table S5).
On the other hand, in EBV+ cases, a worse PFS was found in children with granulomas (p = 0.035, log-rank), FOXP3+ lymphocytes > 50th percentile (p = 0.046, log-rank), FOXP3/CD8 ratio >1 (p = 0.021, log-rank), FOXP3/TIA1 ratio >1 (p = 0.06, log-rank) and TBET/CMAF ratio >1 (p = 0.017, log-rank) (Supporting Information Table S4). Because of the small number of children in this group, Cox analysis was not performed.
The tumor microenvironment in cHL has been considered to be a manifestation of host immune reactions to malignant cells.2 In this context, it is important to note that as yet all cHL tumor microenvironment studies were performed with adults or a mix of adolescents and adults and that it is not clear if children and adolescents have a particular composition of tumor microenvironment. This is a pertinent point because immunological differences exist between children and adults.10, 11 It is conceivable that such differences could affect the immunological response to HRS cells or to viral antigens in EBV+ cases.
In our pediatric group, we have observed higher numbers of CD3+ over CD20+ lymphocytes in the tumor microenvironment, in line with previous studies. However, an analysis by T cell populations showed that B lymphocytes were more frequent than any of the T cell subsets investigated, including CD4+ cells. Thus, this first study of pediatric cHL suggests that the tumor microenvironment composition is indeed different from that observed and adults with cHL, where CD4+ cells are more prevalent than CD20+ lymphocytes2, 4, 30 and that the described predominance of infiltrating CD4+ T cells3, 5, 8 is not an invariable attribute of cHL.
This distinct feature of pediatric cHL may reflect the higher number of CD20+ in relation to CD4+ lymphocytes observed in peripheral blood of healthy children, when compared to adults.12, 13, 21 The quantitative contribution of CD20+ lymphocytes to the cHL microenvironment is starting to attract attention, since this lymphocyte subset has been associated with a favorable response to therapy.18, 31 In our study, the CD20+ population showed a marginal impact on overall survival. A more detailed characterization of intratumoral B-cells, for example with reference to B-1 and B-2 cells, will be required to fully appreciate any effects of B-cells on outcome in pediatric cHL.
It is possible that the shift from a cytotoxic/Th1 tumor microenvironment to a more regulatory/Th2 profile, as observed in older children in our study, may mirror the physiological behavior of lymphocyte populations in the periphery. In line with this, Faria et al. showed that age-related changes in peripheral lymphocyte subsets in healthy Brazilian populations are not progressive and equally steady for all cell populations, but instead show an oscillatory behavior which is characteristic for each cell population in different age groups.21 Previous studies have emphasized the role of cytokines and chemokines produced by the neoplastic cells in shaping the microenvironment of cHL. Our results now raise the possibility that physiological and age-related changes of the immune system may also modulate the tumor microenvironment in pediatric cHL.
In our cases, an intense T cell infiltrate was associated with favorable disease presentation (low disease stages and tumor burden). In colorectal cancer it is well recognized that high numbers of intratumoral CD3+ lymphocytes are associated with a less aggressive disease course.32, 33 In cHL such an association has not been confirmed and may reflect the presence of a systemic mechanism of immunosurveillance.
In this pediatric series, we were not able to confirm the value of FOXP3+ Treg population for prediction of favorable outcome. As in adults,5–7 higher numbers of cytotoxic cells were independently associated with worse PFS.
The positive correlation between the number of benign Ki67+ lymphocytes and different lymphocyte subsets, described for the first time in this study, indicates that lymphocytes from tumor microenvironment are able to proliferate in situ in response to antigens and/or cytokine stimuli. High numbers of Ki67+ lymphocytes showed a borderline association with unfavorable outcome in univariate analysis. This is in apparent contrast to the positive prognostic effect of a high CD3+ T-cell count and suggests that proliferation may be restricted to T-cell subsets with adverse prognostic impact. Additional studies are required to establish the phenotype and functional status of the Ki67+ benign lymphocytes.
This pediatric series was composed of comparable numbers of EBV+ and EBV− cases, allowing an analysis of the role of the EBV in modulating the immune response at the microenvironment level. Although previous in vitro studies showed that EBV can induce the production of cytokines involved in the Treg-CD4+ cell migration,34 we did not observe differences in the number of CD4+, FOXP3+ and CMAF+ cells between EBV+ and EBV− cases. On the other hand, EBV+ cases were characterized by high numbers of intratumoral CD8+, TIA1+, Granzyme B+ and TBET+ cells. This is in line with seminal research that showed that, compared with EBV− cases, EBV+ cHL expresses significantly higher levels of MCH Class I, an intact antigen presentation machinery and higher numbers of intratumoral activated cytotoxic T lymphocytes.35–37 Recently, Chetaille et al.,18 using samples of adults and children, described a molecular signature of EBV+ cHL that is characterized by genes associated with Th1 and antiviral responses, which is validated in our exclusively pediatric population. However, in adults, it is possible that cytotoxic T cells (in parts responsible for this antiviral signature) are not effective in the immune response against EBV in the HRS cells, as previously shown by Frisan et al.29 In addition, acquisition of genetic alterations by HRS cells can result in resistance to apoptosis and therefore resistance to cytotoxic T cell-mediated killing.38–40
In the EBV+ group, characterized by a better therapeutic response, a worse PFS was observed in cases with high number of FOXP3 lymphocytes, FOXP3/CD8 ratio >1 and FOXP3/TIA1 ratio >1 suggesting that: (1) the immune response against EBV is mediated by CD8+ cytotoxic lymphocytes and (2) FOXP3+ lymphocytes are inhibiting the cytotoxic response directed against EBV, as demonstrated by several studies.19, 41, 42 EBV is able to modulate immune escape mechanisms2, 17, 35, 43, 44 and several studies point to an adverse prognosis for EBV-associated adult cHL.45–47 Nevertheless, it appears likely that in pediatric cHL, the interaction between EBV+ HRS and the immune system follows a different scenario. In fact, a favorable outcome was described for EBV+ young cHL patients,48, 49 including the series reported here, suggesting that in children with EBV+ cHL, EBV may elicit an effective immune response, which is negatively modulated by a suppressor microenvironment.42, 50 Cytotoxic T-cells are an adverse factor in EBV− but not in EBV+ cases. On the contrary, FOXP3+ cells had a negative prognostic effect in EBV+ but not in EBV− cases. This became particularly evident when examining the ratio of FOXP3+ Treg over cytotoxic effector cell populations. This reinforces our hypothesis that in pediatric patients, an EBV-specific immune response may be mounted locally and that inhibition of this immune response by FOXP3+ cells may negatively affect outcome. If confirmed, this would have significant implications for the application of microenvironment targeted therapies in the pediatric settings.
We are, nevertheless, aware that the number of cases in this study imposes a limitation in relation to the prognostic impact of the described variables. However, we think our series of sequential cases with similar distribution in relation to EBV-status is appropriate for the immunological evaluation performed here.
In conclusion, this study suggests that the tumor microenvironment of pediatric cHL has particular characteristics, not only with respect to the number of lymphocytes, but also in relation to the functional status of these cells. Validation studies are necessary to confirm the prognostic impact of the variables described in this study.
- 1International Agency for Research on Cancer., World Health Organization. WHO classification of tumours of haematopoietic and lymphoid tissues, 4th edn. Lyon, France: International Agency for Research on Cancer, 2008.,
- 23Principles and practice of pediatric oncology, 3rd edn. Philadelphia: Lippincott-Raven, 1997., .
Additional Supporting Information may be found in the online version of this article.
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.