Comparison of Epstein-Barr Virus Presence in Hodgkin Lymphoma in Pediatric Versus Adult Argentine Patients Open Access

Hodgkin lymphoma is a malignant lymphoma whose unusual heterogeneous clinical, histologic, and epidemiologic features have suggested either a single disease entity with a complex host response or more than one possible distinct etiology.

A definite bimodal age peak, absent in most other lymphomas, is present in the incidence of Hodgkin lymphoma. In developing countries, the early peak occurs in childhood, whereas in developed countries, it occurs in young adults; the second peak occurs in late adulthood in both groups.1,2 Furthermore, in developing countries, there is usually a marked male preponderance and a high proportion of the mixed cellularity histologic subtype. Nodular sclerosis, the predominant histologic subtype in the United States and western Europe, is generally much less common in developing countries.3 

Growing evidence points to an etiologic role for Epstein-Barr virus (EBV), a ubiquitous herpesvirus associated with several lymphoid malignancies and certain epithelial tumors. Early suggestions of EBV and Hodgkin lymphoma association were based on the finding of an altered pattern of serum antibodies to EBV antigens before tumor development.4 Furthermore, a role for EBV in the pathogenesis of Hodgkin lymphoma has been supported by the identification of clonal EBV DNA in a subset of Hodgkin lymphoma on the basis of Southern blot DNA hybridization.5,6 The EBV genome was subsequently localized to Reed-Sternberg cells by in situ hybridization3,7,8; these cells were also shown to express high levels of latent membrane protein 1 (LMP-1) in the absence of EBNA-2.9 Moreover, EBV infection occurs at different ages, depending on the developmental status of the country. In developing countries, EBV infection occurs very early in life. In Argentina, 90% of children have had their first EBV infection before the age of 6 years,10 whereas in developed countries, only 30% to 40% are seropositive by that age.11 

The purpose of this study was to analyze the Hodgkin lymphoma histologic subtype distribution and its association with EBV in different age groups of Argentine patients.

Ninety-two cases of Hodgkin lymphoma in pediatric patients from 1988 to 2001 were selected from the archives of the Service of Pathology at the Ricardo Gutiérrez Children's Hospital (Buenos Aires, Argentina) on the basis of available formalin-fixed, paraffin-embedded biopsies. In addition, a random series of 42 Hodgkin lymphoma cases from the Marie Curie Hospital and another 39 Hodgkin lymphoma cases from the Private Cytology and Pathology Centre (both in Buenos Aires, Argentina) were selected. All were Argentine cases with nodal localization, except for one splenic case. Cases were checked histologically by at least one of us (A.V.B. and/or E.D.M.) and classified according to World Health Organization guidelines.12 

Immunostaining was used to localize LMP-1 expression in tumor cells with monoclonal antibodies CS1, CS2, CS3, and CS4 (Dako Corporation, Carpinteria, Calif). Paraffin sections were deparaffinized and digested with protease XIV (0.1 mg/dL) (Sigma Chemical Company, St Louis, Mo) in phosphate-buffered saline for 15 minutes at 37°C as an antigen unmasking procedure to detect LMP-1.

Each of the nodular lymphocytes predominant Hodgkin lymphoma and classic Hodgkin lymphoma histologic subtypes were immunophenotyped on paraffin-embedded sections of formalin-fixed tissues using antibodies to CD3, CD20, CD30 (Biomeda, Foster City, Calif), CD45 (Dako), CD15 (Becton Dickinson, San José, Calif), and epithelial membrane antigen (Dako). Paraffin sections were deparaffinized and pretreated with a 2.94-g/L (0.01 mol/L) trisodium citrate solution (pH = 6) for 10 minutes at 700 W in a microwave oven.

Endogenous peroxidase activity was blocked with 0.3% (vol/vol) hydrogen peroxide. The immunohistochemical detection of antibodies was performed with a streptavidin-biotin-peroxidase technique (Vectastain Elite ABC, Vector Laboratories, Burlingame, Calif) according to the manufacturer's instructions. Slides were counterstained with hematoxylin and mounted. As a positive control for LMP-1, we used an EBV-positive pediatric nasopharyngeal carcinoma biopsy.

Epstein-Barr encoded RNA in situ hybridization was performed on formalin-fixed, paraffin-embedded tissue sections on 3-aminopropyltriethoxy-silane–treated slides using fluorescein isothiocyanate–conjugated Epstein-Barr encoded RNA oligonucleotides as probes (Novocastra Laboratories, Newcastle, England). Hybridized sites were detected with an alkaline phosphatase–conjugated monoclonal antibody to fluorescein isothyocyanate (Novocastra) according to the manufacturer's instructions. As a positive control for Epstein-Barr encoded RNAs, the same sample as in LMP-1 immunohistochemistry was used.

Data were compared by chi-square and Fisher exact tests when appropriate; P values <.05 were considered significant.13 

The 173 Hodgkin lymphoma cases consisted of 92 children aged 2 through 16 years (median, 8 years) and 81 adults, 39 of whom were from the private center (age range, 17–78 years; median, 28 years) and 42 of whom were from the public hospital (17–83 years; median, 31 years). There was a predominance of males among the pediatric group (male-female ratio, 2:1), whereas in the adult group, the male-female ratio was 1:1. The age distribution of the pediatric population showed 57% (52 of 92) to be younger than 10 years, and the adult group showed a young adult peak (aged ≤35 years) that represented 54% (44 of 81) of the total adult population. However, the Hodgkin lymphoma incidence in the group of patients older than 50 years failed to show the peak previously reported by Glaser et al14 (Table 1).

The histologic-subtype distribution of both groups is shown in Table 2. In the pediatric group, the mixed cellularity prevalence was 48 of 92 (52%), whereas the adult group showed a similar percentage for both nodular sclerosis (38 of 81 [47%]) and mixed cellularity (36 of 81 [44%]) cases that originated from the balance between the histologic-subtype distribution of the private (nodular sclerosis, 22 of 39 [56%]) and public (mixed cellularity, 24 of 42 [57%]) groups.

The detection of EBV was performed by means of Epstein-Barr encoded RNA in situ hybridization. Subcellular localization is a helpful indication of reaction specificity. A nuclear localization of the hybridization signal was detected with a characteristic labeling of the inner nuclear membrane and around the nucleolus of Reed-Sternberg and Hodgkin cells (Figure 1, a and b). In some cases, the intensity of hybridization was so strong that it obscured the nuclear morphology. Among sections from 92 pediatric cases of Hodgkin lymphoma hybridized with Epstein-Barr encoded RNA probes, 51 showed a clear reaction signal in the nuclei of Reed-Sternberg and Hodgkin cells. Nonetheless, only 25 of 81 (31%) of the adult cases were Epstein-Barr encoded RNA positive—16 of 42 (38%) from the public hospital and 9 of 39 (23%) from the private medical center. However, this difference was not statistically significant.

When the Reed-Sternberg and Hodgkin cells from the 92 pediatric cases were challenged with a pool of monoclonal antibodies to LMP-1, the cytoplasm and membranes from the same 51 (55%) cases were strongly labeled (Figure 2). No LMP-1–positive small lymphocytes were detected; however, as previously described by Pallesen et al,9 a weak cross-reactivity with a small fraction of plasma cells was recorded.

In pediatric Hodgkin lymphoma, EBV was significantly associated with the mixed cellularity subtype, since it was present in 77% (37/48) of these cases when compared with the other histologic subtypes of Hodgkin lymphoma (P < .001, χ² test). No difference in the frequency of EBV infection in tumor cells was seen in adult Hodgkin lymphoma when comparing different subtypes (P = .2, χ² test), but EBV was mainly detected in mixed cellularity–subtype cases (14/36 [39%]) (Table 3).

Because, in our country, primary EBV infection occurs mainly in early childhood, EBV positivity in Hodgkin lymphoma biopsies was analyzed separately in 2 pediatric groups younger than 10 years and 11 through 16 years. The incidence of EBV positivity was significantly higher in Hodgkin lymphoma biopsies from patients of the younger age group than from patients of the older age group (75% [39/52] vs 30% [12/40]) (P < .001, the Fisher exact test).

Adult patients with EBV constituted 6 of 44 (14%) patients aged 17 through 35 years, 8 of 15 (53%) patients aged 36 through 50 years, and 11 of 22 (50%) patients aged 51 years and older. The difference among these adult age groups also reached statistical significance (P < .001, χ² test).

It has been known for some time that there is diversity in the incidence, age, and sex distribution and morphology of Hodgkin lymphoma in different populations, and more recently, it has been recognized that the frequency of EBV positivity also varies.

As previously described by Harris,15 there are 3 epidemiologic patterns of Hodgkin lymphoma according to the level of socioeconomic development. In pattern I, seen in undeveloped countries and low socioeconomic groups, Hodgkin lymphoma incidence shows an early childhood peak, no third-decade peak, and a steady rise with advancing age. The predominant histologic subtype is mixed cellularity Hodgkin lymphoma. In pattern III, seen in developed countries and high socioeconomic groups, Hodgkin lymphoma incidence shows a third-decade peak and a predominance of nodular sclerosis over other subtypes of Hodgkin lymphoma. Pattern II is intermediate, with both childhood and second-decade peaks and equal frequencies of mixed cellularity and nodular sclerosis Hodgkin lymphoma subtypes, and is seen in developing or transitional economies. According to such proposed patterns, our study population should be considered pattern II, corresponding to a transitional economy, with a broad peak corresponding to patients younger than 16 years (53% [92/173]) and patients between 17 and 35 years (25% [44/173]) and similar frequencies of nodular sclerosis and mixed cellularity subtypes in the overall population. When adult groups were analyzed separately, nodular sclerosis predominated in the private group according to pattern III, and mixed cellularity predominated in the public group according to pattern I. It is notable that, in our community, pediatric oncologic patients of all socioeconomic groups attend 2 main pediatric hospitals, one of which is our own.

In developing countries, EBV infection occurs very early in life; more than 90% of children are infected by the age of 6 years. Meanwhile, in developed countries, only 30% to 40% are seropositive by that age.11 In Argentina, we have previously shown a prevalence for early childhood EBV infection.10 

We performed LMP-1 immunohistochemistry and Epstein-Barr encoded RNA in situ hybridization to identify cases in which EBV is exclusively expressed in malignant cells. This is normally related to a direct role for the virus in the neoplastic process. Pattern II children who develop Hodgkin lymphoma are more likely to have EBV⁺ Hodgkin lymphoma than are young adult patients. This association of EBV with Hodgkin lymphoma is clearly shown in our study, since 55% of childhood Hodgkin lymphomas were EBV⁺, whereas 14% of young adults (17–35 years) with Hodgkin lymphoma were EBV⁺. In fact, the low incidence of EBV in the young adult group was not merely dependent on the number of studied cases, since 44 patients corresponded to this age group, but only 14% were EBV⁺.

As in our previous report,16 a statistically significant association between EBV-associated Hodgkin lymphoma and younger age was evident (75%) (P < .001, the Fisher exact test). This result is consistent with reports from other developing countries (Honduras,17 Peru,18 Kenya,19 Brazil,20 and China21). Pediatric Hodgkin lymphoma reports from developed countries have also described high rates of EBV infection in children younger than 10 years, but such figures are lower than those from developing countries.14,17,22–24 

The statistically significant association between EBV and the mixed cellularity Hodgkin lymphoma subtype among pediatric patients (P < .001, χ² test) was also described in previous reports from both economically developed and developing countries.14,17,18 Although both nodular sclerosis and mixed cellularity are the most frequent histologic Hodgkin lymphoma subtypes among adult patients, EBV positivity of the mixed cellularity subtype was higher (mixed cellularity = 39% vs nodular sclerosis = 24%), but its association was not statistically significant (P = .2, χ² test). This observation is consistent with other reports, which state that mixed cellularity cases are significantly more likely to be EBV⁺ than are nodular sclerosis cases; positive rates vary from 32% to 96%, with most studies describing rates greater than 50%. By contrast, in nodular sclerosis, this proportion is much lower, ranging from 10% to 50%.25 When we separately analyzed EBV positivity of the mixed cellularity subtype in contrasting socioeconomic groups, we found 11 of 24 (46%) EBV⁺ cases in the public group versus 3 of 12 (25%) EBV⁺ cases in the private group.

In conclusion, our findings further support the suggestion that Hodgkin lymphoma is etiologically heterogeneous. They strengthen the argument that EBV is involved in the pathogenesis of most Hodgkin lymphoma cases in children younger than 10 years and represent a rare complication of early EBV infection. Our recorded EBV prevalences (44% [76/173]), together with both childhood and second-decade peaks as well as similar frequencies of mixed cellularity (48%) and nodular sclerosis (39%), distinguish our population from others in developing countries.

This study was supported in part by a grant from the Ministerio de Salud, Secretaría de Ciencia y Tecnología “Beca de Investigación Ramón Carrillo-Arturo Oñativia.” Mr Chabay was supported by a fellowship from the National Research Council (CONICET), and Dr Preciado is a member of the CONICET Research Career Programme.