Association between Epstein-Barr virus and Thymic epithelial tumors: a systematic review
Infectious Agents and Cancer volume 14, Article number: 32 (2019)
The possible role of Epstein-Barr virus (EBV) in the pathogenesis of thymic epithelial tumors (TET) remains controversial. This study aimed to determine the prevalence of EBV in TET. We conducted a systematic review of relevant English-language studies published between January 1980 and December 2013. Effect size was calculated as event rates (95% confidence interval [CI]) by homogeneity testing using Cochran’s Q and I2 statistics for benign TET, benign TET with myasthenia gravis (MG), and thymic carcinoma (TC). Among 136 potentially relevant studies, 22 met the inclusion criteria. Despite a considerable degree of heterogeneity, the pooled estimated incidences were 9% (95% CI, 1–23%), 20% (95% CI, 0–54%), and 6% (95% CI, 0–21%) for benign TET, benign TET with MG, and TC, respectively. There was significant heterogeneity among studies that used in situ hybridization (ISH) for both benign TET and benign TET with MG. According to the random-effects model, studies employing ISH yielded lower point estimates of EBV prevalence (5%) than those employing other methods (33%). Using the random-effects model, we found a lack of significant heterogeneity among studies from different geographic regions (p = 0.0848). Further, 12 of 23 lymphoepithelioma-like carcinoma (LELC) cases tested EBV-positive. The prevalence of EBV in benign TET with or without MG was lower than in nasopharyngeal carcinoma, suggesting that EBV plays a minor role in TET pathogenesis. Although the prevalence of EBV in TC was also low, EBV may play an important causal role in LELC. Further research is needed to clarify these associations.
Thymic epithelial tumors (TET) are the most common type of primary tumor in the anterior mediastinum, although they are rare. They can be benign (e.g., thymomas, thymic cysts, thymic hyperplasia, and thymolipomas) or malignant (e.g., thymic carcinomas [TCs] and thymic carcinoids). Moreover, approximately 30 to 50% of patients with thymomas have myasthenia gravis (MG), and approximately 80% patients with MG have thymic abnormalities, including thymoma and hyperplasia . Although the mechanism underlying TET is not known, genetic and environmental factors are both suspected to be involved in the etiology of TET [2, 3]. Additionally, infection by common viruses is considered to be a possible environmental factor leading to autoimmunity via alteration of the host immune system.
The Epstein-Barr virus (EBV) is well known for its association with several epithelial neoplasms, inclusive of nasopharyngeal carcinoma, gastric carcinoma, and lymphoepithelioma-like carcinoma (LELC) of several sites (e.g., stomach, salivary gland, and lung) . Next to the direct involvement in the pathogenesis of infectious mononucleosis, EBV is associated with a wide variety of neoplastic lymphoproliferative disorders, including Burkitt lymphoma, Hodgkin disease, post-transplantation lymphoproliferative disease, and non-Hodgkin’s lymphoma arising in immunocompromised patients, as shown in seroepidemiological, immunological, and DNA studies .
Recently, discussion of the possible role of EBV in TET pathogenesis has been stimulated by a study reporting EBER-positive cells in all 17 MG thymic specimens analyzed . Many thymic cells showed positive staining for the EBV-encoded nuclear antigen and latent membrane proteins 1 and 2A. A follow-up study suggested that EBV infection may be involved in the onset or perpetuation of MG . These results, however, could not be reproduced by other groups and thus remain controversial. Two subsequent independent studies did not find evidence of EBV infection in thymic specimens of patients with MG [8, 9].
Since then, several new reports have been published, but a consistent conclusion has not yet been reached. To fill this gap, we conducted a quantitative systematic review of the accumulated evidence on EBV in TET in the published literature, without any restrictions concerning the EBV detection method or geographic origin of the study.
Materials and methods
Literature search and selection criteria
This systematic review was conducted according to the Meta-analysis of Observational Studies in Epidemiology guidelines . A comprehensive search of PubMed, Embase, Ovid, and Google Scholar was performed to identify studies that assessed the presence of EBV infection in human subjects with TET with or without MG published between January 1, 1980 and September 31, 2019. The reference lists of relevant papers and reviews were examined to identify further articles. The key search terms used were Epstein-Barr virus, EBV, thymic, thymus, myasthenia gravis, MG, and thymomas.
Only papers published in English were considered. The included studies provided the exact numbers of analyzed cases and cases testing EBV-positive, which were necessary for calculation of the event rates (EBV prevalence) and their 95% confidence intervals (95% CI). In addition, the included studies provided a clear description of the EBV detection methods and a definite histopathological diagnosis of the thymus. However, studies performed on human cell lines and those with an unknown pathology of the thymus were excluded from the systematic review. Studies with a sample size < 5 cases were also excluded except for those evaluating patients with LELC because of its rarity. Additionally, studies that used enzyme-linked immunoassays for EBV detection and those that evaluated blood samples without detection in thymic tissue were also excluded from the systematic review.
The studies were evaluated critically by two authors (ZG and LC). If disagreement persisted after the complete manuscript was studied, a third reviewer was consulted. From the summary and/or body of the text of each eligible study, we obtained the following information: author of the publication, geographic region of the study, type of specimen, histological type of thymoma analyzed, EBV detection method, total number of cases analyzed, number of cases testing EBV-positive, and percent EBV positivity.
The analysis was conducted using R version 3.6.1 for Windows. No prospective studies that reported EBV status in both TET cases and controls were available. Therefore, odds ratios could not be calculated in an appropriate manner. The proportion of EBV-positive cases was analyzed in 3 groups: benign TET, benign TET with MG, and TC.
We estimated the prevalence using the double arcsine transformation because the inverse variance weight in fixed effects (FE) is suboptimum when dealing with binary data with low prevalence. Additionally, the transformed prevalence was weighted very slightly towards 50%, and studies with a prevalence of zero could thus be included in the analysis [11, 12]. The Q statistics were calculated to test for heterogeneity among the studies included in this analysis and were considered statistically significant when p < 0.05. The I2 statistic was also used to assess the extent of between-study heterogeneity. The DerSimonian and Laird (random-effects) method was selected when there was evidence of statistical heterogeneity. The robustness of the pooled proportions was explored through sensitivity analyses. Sources of heterogeneity were explored using subgroup and meta-regression analyses. Specifically, we chose to study factors related to the EBV detection method and geographic origin.
Our search, which was performed on October 2, 2019, identified 136 articles. Of these articles, 115 were excluded on the basis of their titles or abstracts, and 21 were selected for full-text review. A manual review of the references of relevant publications was also performed to identify additional studies, which yielded 2 additional articles. Ultimately, 22 studies fulfilled the inclusion and exclusion criteria and were included in this review; 15 were larger series related to the prevalence of EBV in thymic tissues, and 7 were case reports on EBV in LELC. Table 1 summarizes the detection of EBV in TET using different detection techniques.
Prevalence of EBV in benign TET
To date, 15 published studies have analyzed the evidence for EBV involvement in benign TET. The reported incidence of EBV in benign TET varied markedly across individual studies; 10 studies reported a prevalence of 0% and one study reported a prevalence of 100%. As significant heterogeneity was observed across the included studies, measured by Cochran’s Q statistics with p < 0.0001, we used a random-effects (RE) model to pool the data. Therefore, the crude EBV positivity (40/482) translated to an EBV prevalence of 0.09 (95% CI 0.01–0.23) using the RE model (Fig. 1).
We hypothesized that appropriate selection of the detection method is important for accurate estimation; in situ hybridization (ISH) is the gold standard for EBV detection, most of the studies used ISH. However, significant heterogeneity was still observed among studies using ISH (p < 0.001, Table 2). ISH gave lower point estimates of EBV prevalence (5%) than other methods (33%) when using the RE model. When the data were analyzed according to regional group, the between-strata comparison using the RE model was not significant at p = 0.0848, indicating that studies from different geographic regions were not significantly different. The incidence was 5% (95% CI, 0–20%) for Asian cases, 19% (95% CI, 0–56%) for European cases, and 0% (95% CI, 0–3%) for other cases (Table 3).
Although there was some evidence of publication bias (Egger’s p = 0.023) there was no evidence for publication bias among studies (n = 12) that used the ISH (Egger’s p = 0.078). Sensitivity analyses were conducted to explore the robustness of these observations, and the results seemed robust to all one-by-one removals of studies, with no change in the magnitude and precision of the FE and RE summary point estimates of the effect size.
Prevalence of EBV in MG
In total, 8 studies analyzed the association between EBV and MG, irrespective of the pathology of TET. Significant heterogeneity was observed among the studies, as measured by Cochran’s Q statistics, with p < 0.001. The crude EBV positivity (36/179) translated to an EBV prevalence of 0.20 (95% CI 0–54%) using the RE model (Fig. 2). Among the 8 studies, the incidence of EBV in MG varied. That is, 4 studies reported no association between EBV and MG, but one study in Caucasian patients reported 100% positivity for EBV.
When the studies were stratified according to EBV detection technique, the between-strata comparison using the RE model was not significant at p = 0.5233, indicating that there was no difference among studies using different EBV detection techniques (Table 4). However, the detection rates tended to be lower in studies that used ISH, although even then a wide variation in detection rate (from 0 to 100%) was evident. The number of studies was too low to draw definite conclusions on the role of EBV in the etiology of MG when the studies were stratified by geographic origin.
Prevalence of EBV in TC
Six studies reported an association between EBV and TC, and all of them used ISH. Significant heterogeneity was observed between the studies, as measured by Cochran’s Q statistics, with p = 0.0026. The crude EBV positivity (9/117) translated to an EBV prevalence of 0.06 (95% CI 0–0.21) using the RE model (Fig. 3). Of the 6 studies, 2 reported no association between EBV and TC, and 3 reported only 1 positive case.
LELC is an undifferentiated or poorly differentiated squamous cell carcinoma associated with a prominent component of reactive lymphocytes and plasma cells. In the English literature, we found a total of 23 cases in which the association between EBV and LELC was investigated; 12 of these cases were positive for EBV, including 8 men and 4 women with a mean (± standard deviation) age of 26.75 ± 19.4 years (range: 10 to 73 years) (Table 5). Therefore, although the EBV positivity was very low in TC, EBV likely contributes to the development of thymic LELC.
The role of EBV in the etiology of TET has attracted increasing interest since the publication of several contradictory reports on the incidence of EBV in TET [6,7,8,9]. However, until now, no systematic review has been published. The present study was a systematic review of published studies on the role of EBV in the etiology of TET. The studies analyzed benign TET, benign TET with MG, and TC. Moreover, the study-level covariates considered in this meta-analysis were the EBV detection method and the geographic origin of the study. Although there was a considerable degree of heterogeneity, the pooled estimated incidences of all 3 categories were 9% (95% CI, 1–23%), 20% (95% CI, 0–54%), and 6% (95% CI, 0–21%) in benign TET, benign TET with MG, and TC, respectively. In addition, studies that evaluated the association between EBV and LELC were also analyzed: 12 of the 23 LELC cases were EBV-positive.
Heterogeneity across study estimates was an important factor limiting the interpretation of our results. Although we attempted to minimize heterogeneity in this review by grouping studies according to the EBV detection method, marked heterogeneity existed in studies of benign TET with and without MG, as assessed by the Q test and I2 index. Although ISH detection of EBERs is a sensitive and specific method for in situ detection of EBV infection in routinely processed tissues and ISH-based studies comprised the bulk of the studies in the present review (12/15 studies), the RE model suggests a lack of true heterogeneity between the studies using different EBV detection techniques, as indicated by the non-significant p-value for homogeneity (p = 0.1931) in the between-study summary comparison.
An alternative view suggests that the variability in EBV prevalence may be related to the different geographic regions of the studies. To test this hypothesis, we performed our analyses stratified according to the geographic origin of the studies. However, when the RE model was applied to calculate the summary statistics, a non-significant (p = 0.0848) summary homogeneity p-value was obtained for the between-strata comparison, which suggests that the variation in EBV prevalence between different geographic regions is not statistically significant. It should also be emphasized, however, that relatively few studies have been published from other geographic regions.
Some evidence for publication bias was detected, but publication bias usually had an insignificant effect on the adjusted point estimates in the stratified meta-analysis. Importantly, there was no evidence of publication bias among the ISH-based studies. As the number of studies was too low for analysis when the studies were stratified by geographic origin, we could not evaluate potential publication bias among studies of different geographic origins. In the sensitivity analysis, all of the meta-analytic results seemed robust to all one-by-one removals of studies, with no change in the magnitude and precision of the FE and RE summary point estimates of the effect size. This finding suggests that none of the 15 studies was sufficiently influential to affect the summary point estimates of this meta-analysis.
Strengths and weaknesses of this review
By synthesizing all of the available published data on thymomas, we have provided a more precise estimate of the prevalence of EBV in TET than was previously available. The influence of variation in study design was diluted by pooling all of the available data, and we were able to explore this heterogeneity by examining study characteristics and conducting subgroup analyses. However, our study also had some limitations. First, some of the studies were not originally designed to determine the prevalence of EBV in TET, and the unknown pathology of TET resulted in incomplete case assessment. Second, the number of studies was low; when the studies were stratified by geographic origin, the number of studies was too low to draw definite conclusions on the role of EBV in the etiology of MG. Furthermore, the published studies only represent certain geographic regions and thus did not provide information on the global status of EBV in TET. Finally, similar to most meta-analytic reviews, we pooled studies that exhibited significant heterogeneity (i.e. in the finding of EBER in TET). In order to reach conclusive evidence on this topic, it may be helpful if international experts in the field of EBV and also of TET could review the published samples.
We have provided evidence regarding the prevalence of EBV in TET. Compared to the necessary role of EBV in nasopharyngeal carcinoma, which has been well documented in the literature, the prevalence of EBV in benign TET with or without MG was low, which suggests that EBV plays a minor role in the pathogenesis of TET. In the analyses stratified according to EBV detection method and geographic origin of the study, the variability among studies did not reach statistical significance when RE models were used. However, more data are necessary to better evaluate the impact of EBV in the pathogenesis of TET. Although the prevalence of EBV in TC is also low, EBV may play an important causal role in LELC. Further research is necessary to shed light on these associations.
Availability of data and materials
Le Panse R, Bismuth J, Cizeron-Clairac G, Weiss JM, Cufi P, Dartevelle P, et al. Thymic remodeling associated with hyperplasia in myasthenia gravis. Autoimmunity. 2010;43:401–12. https://doi.org/10.3109/08916930903563491.
Klavinskis LS, Willcox HN, Richmond JE, Newsom-Davis J. Attempted isolation of viruses from myasthenia gravis thymus. J Neuroimmunol. 1986;11:287–99. https://doi.org/10.1016/0165-5728(86)90082-2.
Smith CI, von Sydow M, Hammarstrom L. Viral etiology in myasthenia gravis? Analysis of thymic tissue. Eur Neurol. 1986;25:317–9.
Niedobitek G, Meru N, Delecluse HJ. Epstein-Barr virus infection and human malignancies. Int J Exp Pathol. 2001;82:149–70. https://doi.org/10.1111/j.1365-2613.2001.iep190.x.
Cohen JI. Epstein-Barr virus infection. N Engl J med. 2000;343:481–92. https://doi.org/10.1056/NEJM200008173430707.
Cavalcante P, Serafini B, Rosicarelli B, Maggi L, Barberis M, Antozzi C, et al. Epstein-Barr virus persistence and reactivation in myasthenia gravis thymus. Ann Neurol. 2010;67:726–38. https://doi.org/10.1002/ana.21902.
Cavalcante P, Maggi L, Colleoni L, Caldara R, Motta T, Giardina C, et al. Inflammation and epstein-barr virus infection are common features of myasthenia gravis thymus: possible roles in pathogenesis. Autoimmune Dis. 2011;2011:213092. https://doi.org/10.4061/2011/213092.
Meyer M, Hols AK, Liersch B, Leistner R, Gellert K, Schalke B, et al. Lack of evidence for Epstein-Barr virus infection in myasthenia gravis thymus. Ann Neurol. 2011;70:515–8. https://doi.org/10.1002/ana.22522.
Kakalacheva K, Maurer MA, Tackenberg B, Munz C, Willcox N, Lunemann JD. Intrathymic Epstein-Barr virus infection is not a prominent feature of myasthenia gravis. Ann Neurol. 2011;70:508–14. https://doi.org/10.1002/ana.22488.
Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of observational studies in epidemiology (MOOSE) group. JAMA. 2000;283:2008–12. https://doi.org/10.1001/jama.283.15.2008.
Beijer U, Wolf A, Fazel S. Prevalence of tuberculosis, hepatitis C virus, and HIV in homeless people: a systematic review and meta-analysis. Lancet Infect Dis. 2012;12:859–70. https://doi.org/10.1016/S1473-3099(12)70177-9.
Freeman MF, Tukey JW. Transformations related to the angular and the square root. Ann Math Stat. 1950;21:607–11. https://doi.org/10.1214/aoms/1177729756.
McGuire LJ, Huang DP, Teoh R, Arnold M, Wong K, Lee JC. Epstein-Barr virus genome in thymoma and thymic lymphoid hyperplasia. Am J Pathol. 1988;131:385–90.
Inghirami G, Chilosi M, Knowles DM. Western Thymomas lack Epstein-Barr-virus by southern blotting analysis and by polymerase chain-reaction. Am J Pathol. 1990;136:1429–36.
Borisch B, Kirchner T, Marx A, Mullerhermelink HK. Absence of the Epstein-Barr-virus genome in the Normal Thymus, Thymic epithelial tumors, Thymic lymphoid hyperplasia in a European population. Virchows Arch B Cell Pathol Incl Mol Pathol. 1990;59:359–65. https://doi.org/10.1007/BF02899425.
Mann RB, Wu TC, MacMahon EM, Ling Y, Charache P, Ambinder RF. In situ localization of Epstein-Barr virus in thymic carcinoma. Mod Pathol. 1992;5:363–6.
Wu TC, Kuo TT. Study of Epstein-Barr virus early RNA 1 (EBER1) expression by in situ hybridization in thymic epithelial tumors of Chinese patients in Taiwan. Hum Pathol. 1993;24:235–8. https://doi.org/10.1016/0046-8177(93)90031-B.
Fujii T, Kawai T, Saito K, Fukushima K, Hasegawa T, Tokunaga M, et al. EBER-1 expression in thymic carcinoma. Acta Pathol Jpn. 1993;43:107–10. https://doi.org/10.1111/j.1440-1827.1993.tb01118.x.
Hishima T, Fukayama M, Hayashi Y, Fujii T, Ooba T, Funata N, et al. CD70 expression in thymic carcinoma. Am J Surg Pathol. 2000;24:742–6. https://doi.org/10.1097/00000478-200005000-00014.
Engel PJH. Absence of latent Epstein-Barr virus in thymic epithelial tumors as demonstrated by Epstein-Barr-encoded RNA (EBER) in situ hybridization. Apmis. 2000;108:393–7. https://doi.org/10.1034/j.1600-0463.2000.d01-74.x.
Tateyama H, Saito Y, Fujii Y, Okumura M, Nakamura K, Tada H, et al. The spectrum of micronodular thymic epithelial tumours with lymphoid B-cell hyperplasia. Histopathology. 2001;38:519–27. https://doi.org/10.1046/j.1365-2559.2001.01133.x.
Chen PC, Pan CC, Yang AH, Wang LS, Chiang H. Detection of Epstein-Barr virus genome within thymic epithelial tumours in Taiwanese patients by nested PCR, PCR in situ hybridization, and RNA in situ hybridization. J Pathol. 2002;197:684–8. https://doi.org/10.1002/path.1141.
Jing F, Wei D, Wang D, Li N, Cui F, Yang F, et al. Lack of Epstein-Barr virus infection in Chinese myasthenia gravis patients. Acta Neurol Scand. 2013;128:345–50. https://doi.org/10.1111/ane.12124.
Leyvraz S, Henle W, Chahinian AP, Perlmann C, Klein G, Gordon RE, et al. Association of Epstein-Barr virus with thymic carcinoma. N Engl J Med. 1985;312:1296–9. https://doi.org/10.1056/NEJM198505163122006.
Dimery IW, Jin SL, Blick M, Pearson G, Spitzer G, Hong WK. Association of the Epstein-Barr virus with Lymphoepithelioma of the Thymus. Cancer. 1988;61:2475–80. https://doi.org/10.1002/1097-0142(19880615)61:12<2475::AID-CNCR2820611215>3.0.CO;2-5.
Matsuno Y, Mukai K, Uhara H, Akao I, Furuya S, Sato Y, et al. Detection of Epstein-Barr-virus DNA in a Japanese case of Lymphoepithelioma-like Thymic carcinoma. Jpn J Cancer Res. 1992;83:127–30.
Patton DF, Ribeiro RC, Jenkins JJ, Sixbey JW. Thymic carcinoma with a defective Epstein-Barr-virus encoding the Bzlf1 Transactivator. J Infect Dis. 1994;170:7–12. https://doi.org/10.1093/infdis/170.1.7.
Niehues T, Harms D, Jurgens H, Gobel U. Treatment of pediatric malignant thymoma: long-term remission in a 14-year-old boy with EBV-associated thymic carcinoma by aggressive, combined modality treatment. Med Pediatr Oncol. 1996;26:419–24. https://doi.org/10.1002/(SICI)1096-911X(199606)26:6<419::AID-MPO10>3.0.CO;2-6.
Koppula BR, Pipavath S, Lewis DH. Epstein-Barr virus (EBV) associated lymphoepithelioma-like thymic carcinoma associated with paraneoplastic syndrome of polymyositis: a rare tumor with rare association. Clin Nucl Med. 2009;34:686–8. https://doi.org/10.1097/RLU.0b013e3181b53f5a.
Januszkiewicz D, Wziatek A, Nowak J. Is lymphoepithelioma like carcinoma of thymic caused by ebv infection? Int J Infect Dis. 2012;16:E87.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Zhang, G., Yu, Z., Shen, G. et al. Association between Epstein-Barr virus and Thymic epithelial tumors: a systematic review. Infect Agents Cancer 14, 32 (2019). https://doi.org/10.1186/s13027-019-0254-5
- Epstein-Barr virus
- Thymic epithelial tumor
- Thymic carcinomas
- Myasthenia gravis
- Systematic review