EBV is an etiological agent in infectious mononucleosis, is implicated in some malignant lymphomas and lymphoepithelioma-like carcinomas, and has been sporadically reported in carcinomas of the breast, lung, and other sites. We studied immunohistochemically benign and malignant tumors of the breast, lung, colon, and prostate and found EBV in some carcinomas of those sites. Also, EBV reactions were noted in hyperplasias and dysplasias, e.g., breast carcinomas in situ and prostatic intraepithelial neoplasia. Benign tumor counterparts were negative. PCR analysis of selected cases confirmed the presence of EBV. Our results suggest that EBV is not restricted to lymphoepithelioma-like carcinomas but may play an oncogenic role in frequent epithelial cancers and possibly also in hyperplasias and certain dysplasias preceding carcinomas.

EBV is a DNA virus belonging to the herpes family; its port of entry is considered to be the oropharyngeal epithelium (1). CD21 has been established as the receptor for EBV in lymphoid B cells (2). In turn, EBV has been established as the etiological agent of infectious mononucleosis (3). In addition, EBV has been associated with a variety of Hodgkin’s and non-Hodgkin’s lymphomas (4), particularly Burkitt’s lymphoma (Ref. 5 and references therein). EBV is also important in the pathogenesis of LELCs3 of the pharynx and other sites (6, 7). More recently, there have been scattered reports linking EBV with conventional epithelial cancers of other primary sites including breast (8, 9, 10), lung (11, 12, 13), and gastric (14, 15) carcinomas. We undertook a systematic immunohistochemical and molecular study of a series of breast, lung, colon, and prostate carcinomas and found that EBV is demonstrable in a variable proportion of these frequent epithelial malignancies. Moreover, we showed that EBV is demonstrable is some hyperplastic and preneoplastic epithelial proliferations of these organs. These data, combined with the fact that the carcinomas studied herein are exceedingly frequent, suggest that EBV, rather than being associated with restricted groups of uncommon malignancies, may in fact play a wider role in human epithelial carcinogenesis.

Detection of EBNA-1 by Immunohistochemistry.

Tissue samples were obtained from the files of Rush-Presbyterian-St. Luke’s Medical Center, Chicago. These cases have been extensively characterized and reported (Ref. 16 and references therein). Formalin-fixed, paraffin-embedded sections were used to immunolocalize EBNA-1 expression applying the Mab 2B4 (kindly provided by Dr. E. Kremmer, Forschungszentrum für Umwelt und Gesundheit GmbH, Institut für Immunologie, Munich, Germany; Ref. 9), LMP-1 expression using CS1-4 (Dako, Carpinteria, CA), ZEBRA expression using BZ-1, and CD21 expression using Mab 1F8 (Dako). Antigen unmasking was done in 10 mm Tris and 1 mm EDTA solution (pH 9.5) or 0.01m sodium citrate (pH 6) for 5 min at 500 watts in a microwave oven. Immunohistochemical detection of Mabs was performed with a streptavidin-biotin complex-peroxidase system (Vectastain Elite ABC; Vector, Burlingame, CA) according to the manufacturer’s instructions. Hematoxylin counterstaining was applied. As negative controls, the pertinent antibodies were omitted. As positive controls for EBNA-1, we used well-characterized LELCs.

PCR Amplification.

PCR amplification was performed on paraffin sections of selected pertinent cases. To confirm EBV presence, a PCR was carried out by amplifying EBV DNA sequences encoding for EBERs as described by Bonnet et al.(9). In a separate reaction tube, a PCR was done to amplify a β-globin gene as a control to monitor the amplification ability of a single-copy gene. Amplified DNA was subjected to electrophoresis on a 2% agarose gel with ethidium bromide.

EBERs in Situ Hybridization.

In situ hybridization was performed on paraffin sections according to the manufacturer’s instructions using a Peptide Nucleic Acid ISH Detection kit and EBER peptide nucleic acid probe labeled with FITC (Dako).

The most significant immunohistochemical results are summarized in Table 1.

Breast.

In 14 of 33 (42%) carcinomas, convincing nuclear reactions were noted involving a range of 5–30% of neoplastic cells; ductal and lobular variants of carcinoma were similarly involved (Fig. 1, a and b); foci of in situ carcinoma also showed focal nuclear staining (Fig. 1,c). Proliferative variants of fibrocystic disease with variable degrees of atypia, and occasionally a lack thereof, including ductal and lobular hyperplasia and papillomas (Fig. 1 d), also showed focal nuclear immunoreactivity. These changes were found in cases with and without an associated carcinoma. In a cellular fibroadenoma (phyllodes tumor), EBV-reactive nuclei were found in some epithelial and stromal cells. Twenty-one normal breast, nonproliferative variants of fibrocystic changes and benign fibroadenomas were negative. CD21 showed no reaction in any epithelial component. Our demonstration of EBV in breast carcinomas confirm and broaden earlier reports including the relative incidence of positive cases (8, 9, 10). However, our findings on EBV in typical and atypical ductal and lobular proliferations and in situ carcinomas represent novel observations and may reflect a possible etiological role of EBV in breast carcinogenesis. Notably, these observations in the breast, with a different EBV pattern of expression (EBER−, LMP-1−, and EBNA-1+), parallel findings and hypothesis advocated by Pathmanathan et al.(17) in nasopharyngeal carcinomas involving dysplasia or preinvasive carcinoma in situ (EBER+, LMP-1+), both representing evidence that EBV may be a primary etiological agent in a multistep process that leads to the development of a carcinoma.

Lung.

In 4 of 54 (7.4%) carcinomas, EBV nuclear immunostaining was noted involving a range of 5–10% of neoplastic cells. All major histological types of lung carcinoma, i.e., squamous, adenocarcinoma, and neuroendocrine carcinoma, were represented among the positive cases (Fig. 1, e and f). Two cases, although classified as squamous and neuroendocrine carcinomas, respectively, because of the heavy predominance of these phenotypes, showed focal adenocarcinoma. Notably, reactions were recognized in both patterns of differentiation. No EBV reactions were noted in 6 hyperplastic bronchial and alveolar epithelium or normal lung, nor was any staining recognized in any of 12 bronchial carcinoids studied. No CD21 reactions were detected in epithelial cells.

Previous studies of EBV in lung carcinomas showed the presence of the pertinent antigen in variably differentiated squamous and adenocarcinomas (11, 12, 13) and specifically mentioned its absence in neuroendocrine carcinomas (11). In this context, our results differ because we showed convincingly EBV in cases of pulmonary neuroendocrine carcinomas with variable degrees of differentiation. These findings suggest that EBV may play a role in lung carcinogenesis regardless of differentiation lineage. Moreover, the consistent absence of EBV in bronchial carcinoids underscores the prevalent notion that these benign neuroendocrine tumors are neither precursors of, nor are they associated with, pulmonary neuroendocrine malignancies.

Colon.

A single case of 19 carcinomas showed convincing EBV reactivity in an estimated 5% of nuclei (Fig. 1,g). Similarly, only a single case of hyperplastic colonic mucosa showed occasional positive nuclei. Paradoxically these cells were topographically close to a colloid carcinoma, which proved to be EBV negative. No epithelial cells showed CD21 staining, whereas B lymphocytes in the vicinity were strongly reactive (Fig. 1 h). Four normal colonic mucosa and 8 benign polyps were negative. As indicated above, we found strong and extensive EBV reactions in a single case of a moderately differentiated colonic adenocarcinoma. Notably, tubular and villous adenomas in this and other cases proved to be nonreactive. Conversely, in the vicinity of a nonreactive colloid carcinoma, the hyperplastic colonic mucosa displayed scattered EBV-positive cells. The numbers are admittedly small, and we can offer no satisfactory explanation for these apparently contradictory findings except that they may underscore the latency and possible opportunism of EBV which, in diverse circumstances, may or may not play an oncogenic role.

Prostate.

Seven of 19 carcinomas showed strong EBV reactions in 5–30% of neoplastic nuclei. These cases included all grades of the Gleason classification ranging from well-differentiated adenocarcinomas (Fig. 1,i) to hypernephroid and solid tumors. One of 7 carcinomas included foci of high-grade PIN, where convincing EBV reactive nuclei were seen in both basal and luminal cells (Fig. 1 j). All prostate carcinoma samples included foci of benign glandular hyperplasia, 3 of which showed sporadic positive nuclei; 10 normal prostate samples were negative. CD21 was not detected in epithelial cells. Our findings of EBV in prostatic carcinomas are entirely new. In these cases, the presence and possibility of EBV appeared unrelated to the degree of neoplastic differentiation because it was found in the entire range of tumors from those composed of regular glands to poorly differentiated and hypernephroid carcinomas. Interestingly, an architecturally and cytologically distorted carcinoma removed after androgen deprivation therapy showed abundant EBV-reactive cells. Also, in a single case of carcinoma, foci of high-grade PIN, acknowledged to be a precancerous proliferation, showed EBV immunostaining. This apparently low incidence may simply reflect the fact that the blocks containing carcinoma selected for this study were not chosen for the concomitant presence of PIN. Perhaps surprising was the frequent finding of scattered immunoreactive EBV cells in benign glandular hyperplasia. These results suggest that EBV may play a role in clearly proliferative but not necessarily malignant or premalignant lesions. In this context, our findings in the prostate parallel the above-described findings in nonprecancerous epithelial proliferations of the breast and colon. Conversely, the presence of EBV in dysplastic and precancerous proliferations of the prostate and breast underscores that EBV may indeed have an optional role in the development of carcinomas of these sites.

Selected positive samples were marked with anti-BZLF1 and with anti-LMP-1; the results were negative, meaning no active replication of EBV in the former case and showing a different latency expression pattern from others previously described for other malignancies in the latter. This is in accordance with the negative results obtained for the same antigens by Chu et al.(18). Certain positive samples were also tested by EBER in situ hybridization, and all cases were negative. This would seem to fit with the negative results obtained with the same technique in a series of 107 breast carcinomas reported by Glaser et al.(19) and in a smaller series reported by Bonnet et al.(9).

By PCR, the EBV genome was detected in 14 of 14 breast carcinomas, 2 of 2 lung carcinomas, and 3 of 4 prostate carcinomas (Table 2; Fig. 2). The EBV genome was not detected in any of the 15 immunohistochemically EBV-negative carcinomas and normal controls.

General Overview.

In the EBV envelope, there are four significant glycoproteins, of which the one known as gp350 interacts with the CD21 receptor present on the surface of B lymphocytes. Thus, the virus enters the latter cells. The remaining glycoproteins are thought to be accessories to the process of viral entry into the cells. In our study, we attempted to demonstrate CD21 in hyperplastic, dysplastic, and transformed epithelial cells but failed to detect it in all instances, whereas CD21 was readily demonstrated in neighboring B lymphocytes. These findings parallel a recent report of CD21-negative, EBV-associated leiomyosarcomas (20). We therefore suggest that the entry of EBV into these epithelial cells may be mediated by either another receptor or by a different mechanism that may not require a specific receptor.

Our immunohistochemical and molecular findings of EBV in a certain proportion of carcinomas of various sites and also in some premalignant and nonpremalignant epithelial proliferations of the same sites suggest that the relationship between EBV and these lesions is neither linear nor simple. We speculate that the significance of EBV in epithelial proliferations may range from that of simple commensal to that of trigger or cofactor in carcinogenesis. In this context, we may add that parallel studies in sequential biopsies of epithelial hyperplasias and dysplasias known or suspected to be precursors of carcinoma in sites such as breast, colon, and prostate may shed light on the possible role that EBV may play in the development of some epithelial cancers.

We thank Anne-Marie Fornabaro for outstanding clerical and secretarial help.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This study was supported in part by a grant from Ministerio de Salud, Secretaria de Ciencia y Tecnologia “Beca de Investigación Ramón Carrillo-Arturo Oflativia” and by a Thrid World Academy of Sciences (TWAS) grant no. 00-232/RG/BIO/LA and a generous gift from the Barry Pitler Fund. P.A.C. was supported by a fellowship from the National Research Council (CONICET) and M.V.P. is a member of the CONICET Research Cancer Programme.

3

The abbreviations used are: LELC, lymphoepithelioma-like carcinoma; Mab, monoclonal antibody; EBER, EBV-encoded early RNA; PIN, prostatic intraepithelial neoplasia.

Fig. 1.

a, breast: infiltrating ductal carcinoma. Note abundant nuclei reactive with Mab 2B4. ×260. b, breast: infiltrating lobular carcinoma; scattered nuclei reactive with Mab 2B4 are evident (arrows). ×440. c, breast: lobular carcinoma in situ. Most neoplastic nuclei comprising the solid and expanded acini reacted with MAb 2B4, albeit with variable intensity (arrows). ×660. d, breast: portion of papilloma without atypia showing nuclei reactive with Mab 2B4 (arrows). ×260. e, lung: squamous carcinoma; note minority of nuclei reactive with Mab 2B4 (arrows). ×260. f, Lung: well-differentiated neuroendocrine carcinoma. Note typically solid pattern, lack of necrosis, and numerous nuclei reactive with Mab 2B4 (arrows). ×260. g, colon: moderately differentiated adenocarcinoma; note cribriform pattern and focally prominent nuclei immunostained with Mab 2B4. ×260. h, colon: colloid variant of adenocarcinoma (Ca), which proved negative with Mab 2B4 (not shown). Note rich CD21 reaction of B lymphocytes in lymphoid aggregate (L) in the vicinity of the tumor, whereas tumor nuclei are unstained. ×260. i, prostate: well-differentiated adenocarcinoma. Abundant nuclei reactive with Mab 2B4 are evident. ×260. j, prostate: high grade PIN. Note nuclei reactive with Mab 2B4 in basal cells (short arrow) as well as in luminal cells (long arrow). ×220.

Fig. 1.

a, breast: infiltrating ductal carcinoma. Note abundant nuclei reactive with Mab 2B4. ×260. b, breast: infiltrating lobular carcinoma; scattered nuclei reactive with Mab 2B4 are evident (arrows). ×440. c, breast: lobular carcinoma in situ. Most neoplastic nuclei comprising the solid and expanded acini reacted with MAb 2B4, albeit with variable intensity (arrows). ×660. d, breast: portion of papilloma without atypia showing nuclei reactive with Mab 2B4 (arrows). ×260. e, lung: squamous carcinoma; note minority of nuclei reactive with Mab 2B4 (arrows). ×260. f, Lung: well-differentiated neuroendocrine carcinoma. Note typically solid pattern, lack of necrosis, and numerous nuclei reactive with Mab 2B4 (arrows). ×260. g, colon: moderately differentiated adenocarcinoma; note cribriform pattern and focally prominent nuclei immunostained with Mab 2B4. ×260. h, colon: colloid variant of adenocarcinoma (Ca), which proved negative with Mab 2B4 (not shown). Note rich CD21 reaction of B lymphocytes in lymphoid aggregate (L) in the vicinity of the tumor, whereas tumor nuclei are unstained. ×260. i, prostate: well-differentiated adenocarcinoma. Abundant nuclei reactive with Mab 2B4 are evident. ×260. j, prostate: high grade PIN. Note nuclei reactive with Mab 2B4 in basal cells (short arrow) as well as in luminal cells (long arrow). ×220.

Close modal
Fig. 2.

PCR performed on paraffin sections of pertinent cases. Cellular eluates were DNA-β-globin positive before PCR was performed. Lanes 1 and 2, breast carcinomas; Lane 3, prostate carcinomas; Lane 4, a positive control of a known, positive cell line (EBV-positive lymphoblastoid cell line P3HR1); Lane 5, a negative control (distilled H2O replacing DNA); Lane 6, a molecular size marker.

Fig. 2.

PCR performed on paraffin sections of pertinent cases. Cellular eluates were DNA-β-globin positive before PCR was performed. Lanes 1 and 2, breast carcinomas; Lane 3, prostate carcinomas; Lane 4, a positive control of a known, positive cell line (EBV-positive lymphoblastoid cell line P3HR1); Lane 5, a negative control (distilled H2O replacing DNA); Lane 6, a molecular size marker.

Close modal
Table 1

Carcinomas, benign tumors, and tumor-like conditions studied by immunohistochemistry for EBV

EBNA-1+%+
Breast  42% 
 Infiltrating ductal carcinoma 8/20  
 Infiltrating lobular carcinoma 6/13  
 Fibroadenoma 1/10a  
 Fibrocystic changes 2/16  
Lung  7.4% 
 Squamous cell carcinoma 2/22  
 Adenocarcinoma 0/22  
 Neuroendocrine carcinoma 2/10  
 Carcinoids 0/12  
Colon   
 Adenocarcinoma 1/19 5% 
Prostate   
 Adenocarcinoma 7/19 37% 
EBNA-1+%+
Breast  42% 
 Infiltrating ductal carcinoma 8/20  
 Infiltrating lobular carcinoma 6/13  
 Fibroadenoma 1/10a  
 Fibrocystic changes 2/16  
Lung  7.4% 
 Squamous cell carcinoma 2/22  
 Adenocarcinoma 0/22  
 Neuroendocrine carcinoma 2/10  
 Carcinoids 0/12  
Colon   
 Adenocarcinoma 1/19 5% 
Prostate   
 Adenocarcinoma 7/19 37% 
a

Positive nuclei in phyllodes tumor variant.

Table 2

Correlation between immunohistochemical results with Mab 2B4 for EBNA-1 and PCR

All samples were conventional paraffin sections. The cases studied with PCR were DNA-β-globin positive.

IHCa EBNA-1+PCR EBV+
Breast 14 14 
Lung 
Prostate 
IHCa EBNA-1+PCR EBV+
Breast 14 14 
Lung 
Prostate 
a

IHC, immunohistochemistry.

1
Epstein M., Achong B., Barr Y. Virus particles in cultured lymphoblasts from Burkitt’s lymphoma.
Lancet
,
1
:
702
-703,  
1964
.
2
Nemerow G., Mold C., Schwend V., Tollefson V., Cooper N. Identification of gp350 as the viral glycoprotein mediating attachment of Epstein Barr virus to the EBV/C3d receptor of B cells: sequence homology of gp350 and c3 complement fragment C3d.
J. Virol.
,
61
:
1416
-1420,  
1987
.
3
Niederman J., McCollum R., Henle G., Henle W. Infectious mononucleosis. Clinical manifestations in relation to EB virus antibodies.
J. Am. Med. Assoc.
,
203
:
205
-209,  
1968
.
4
Preciado M. V., De Matteo E., Diez B., Menarguez J., Grinstein S. Presence of Epstein Barr virus and strain type assignment in Argentine childhood Hodgkin’s disease.
Blood
,
86
:
3922
-3929,  
1995
.
5
Crawford D. Biology and disease associations of Epstein Barr virus.
Phil. Trans. R. Soc. Lond. B
,
356
:
461
-473,  
2001
.
6
Zubizarreta P., D’Antonio G., Raslawski E., Gallo G., Preciado M. V., Casak S., Scopinaro M., Morales G., Sackman-Muriel F. Nasopharyngeal carcinoma in childhood and adolescence. A single institution experience with combined therapy.
Cancer (Phila.)
,
89
:
690
-695,  
2000
.
7
Kieff E. EBV: increasing evidence of a link to carcinoma.
N. Engl. J. Med.
,
333
:
724
-726,  
1995
.
8
Labrecque L., Barnes D., Fentiman I., Griffin B. Epstein Barr virus epithelial cell tumors: a breast cancer study.
Cancer Res.
,
55
:
39
-45,  
1995
.
9
Bonnet M., Guinebretiere J., Kremmer E., Grunewald V., Benhamou E., Contesso G. Detection of Epstein Barr virus in invasive breast cancers.
J. Natl. Cancer Inst.
,
91
:
1376
-1381,  
1999
.
10
Fina F., Romain S., Ouafik L., Palmari J., Ben Ayed F., Benharkat S., Bonnier P., Spyratos F., Foekens J., Rose C., Buisson M., Gerard H., Reymond M., Seigneurin J., Martin P. Frequency and genome load of Epstein Barr virus in 509 breast cancers from different geographical areas.
Br. J. Cancer
,
86
:
783
-790,  
2001
.
11
Wong M., Chung L., Yuen S., Leung S., Chan S., Wang E., Fu K. In situ detection of Epstein-Barr virus in non-small cell lung carcinomas.
J. Pathol.
,
177
:
233
-240,  
1995
.
12
Han A., Xiong M., Gu Y., Lin S., Xiong M. Lymphoepithelioma like carcinoma of the lung with a better prognosis. A clinicopathologic study of 32 cases.
Am. J. Clin. Pathol.
,
115
:
841
-850,  
2001
.
13
Castro C., Ostrowski M., Barrios R., Green L., Popper H., Powell S., Cagle P., Ro J. Relationship between Epstein Barr virus and lymphoepithelioma like carcinoma of the lung: a clinicopathologic study of six cases and review of the literature.
Hum. Pathol.
,
32
:
863
-872,  
2001
.
14
Koriyama C., Akiba S., Iriya K., Yamaguti T., Hamada G., Itoh T., Eizuru Y., Aikou T., Watanabe S., Tsugane S., Tokunaga M. Epstein Barr virus-associated gastric carcinoma in Japanese Brazilians and non-Japanese Brazilians in Sao Paulo.
Jpn. J. Cancer Res.
,
92
:
911
-917,  
2001
.
15
Takada K. Epstein-Barr virus and gastric carcinoma.
Mol. Pathol.
,
53
:
255
-261,  
2000
.
16
Gould V., Martinez N., Orucevic A., Schneider J., Alonso A. A novel, nuclear pore associated molecule overexpressed in oncogenesis and development.
Am. J. Pathol.
,
157
:
1605
-1613,  
2000
.
17
Pathmanathan R., Prasad U., Sadler R., Flynn K., Raab-Traub N. Clonal proliferation of cells infected with Epstein Barr virus in preinvasive lesions related to nasopharyngeal carcinoma.
N. Engl. J. Med.
,
333
:
693
-698,  
1995
.
18
Chu P., Chang K., Chen Y., Chen W., Weiss L. No significant association of Epstein Barr virus infection with invasive breast carcinoma.
Am. J. Pathol.
,
159
:
571
-578,  
2001
.
19
Glaser S., Ambinder R., DiGiuseppe J., Horn-Ross P., Hsu J. Absence of Epstein Barr virus EBER-1 transcripts in an epidemiologically diverse group of breast cancers.
Int. J. Cancer
,
75
:
555
-558,  
1998
.
20
Reyes C., Abuzaitoun O., De Jong A., Hanson C., Langston C. Epstein-Barr virus-associated smooth muscle tumors in ataxia-telangiectasia: a case report and review.
Hum. Pathol.
,
33
:
133
-136,  
2002
.