Epstein-Barr Virus Antibodies and the Risk of Associated Malignancies: Review of the Literature

Transcription

1 American Journal of Epidemiology Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of Public Health This work is written by (a) US Government employee(s) and is in the public domain in the US. Vol. 180, No. 7 DOI: /aje/kwu176 Advance Access publication: August 28, 2014 Review Epstein-Barr Virus Antibodies and the Risk of Associated Malignancies: Review of the Literature Anna E. Coghill* and Allan Hildesheim * Correspondence to Dr. Anna E. Coghill, National Cancer Institute, 9609 Medical Center Drive, Rockville, MD ( anna.coghill@nih.gov). Initially submitted November 27, 2013; accepted for publication June 4, Epstein-Barr virus (EBV), a ubiquitous herpes virus that infects 90% of humans by adulthood, is linked to the development of various cancers, including nasopharyngeal carcinoma, gastric cancer, Burkitt lymphoma, non- Hodgkin lymphoma (NHL), and Hodgkin lymphoma. We reviewed the literature published since 1980 regarding an association between antibodies against EBV proteins and the risk of EBV-associated malignancies. Immunoglobulin A antibody levels that are elevated before diagnosis have consistently been associated with the risk of nasopharyngeal carcinoma, and patients with Hodgkin lymphoma have significantly higher immunoglobulin G antibody levels than disease-free controls. However, the link between the immune response to EBV and other EBV-associated malignancies was less clear. Although evidence of an association between the risk of Burkitt lymphoma and immunoglobulin G antibodies was consistent for available studies, the sample sizes were limited. Evidence for a link between antibodies against EBV and risk of either gastric cancer or NHL was inconsistent. Future investigations should account for tumor EBV status because only 7% 10% of gastric tumors and select NHL subtypes are related to EBV infection. Comparing differences in the associations between the humoral immune response to EBV and disease risk across cancers may help elucidate how this ubiquitous virus contributes to distinct tumors globally. Burkitt lymphoma; EBV antibodies; EBV immune response; EBV serology; gastric carcinoma; Hodgkin lymphoma; nasopharyngeal carcinoma; non-hodgkin lymphoma Abbreviations: BL, Burkitt lymphoma; CI, confidence interval; EA, early antigen; EBNA, EBV nuclear antigen; EBV, Epstein-Barr virus; GC, gastric cancer; HL, Hodgkin lymphoma; IgA, immunoglobulin A; IgG, immunoglobulin G; NHL, non-hodgkin lymphoma; NPC, nasopharyngeal carcinoma; VCA, viral capsid antigen Epstein-Barr virus (EBV), a human herpes virus primarily transferred by oral secretions, persists as a latent infection in human memory B-cells (1 5). Although the average age of childhood infection varies according to geographic region, more than 90% of adults globally are estimated to be infected with EBV (6). In developed countries such as the United States, more than 80% of persons who are tested for EBV are found to be positive by the age of 19 years (7). Malignant transformation of B-cells or epithelial cells infected with EBV can result in the development of various EBV-associated cancers in both children and adults. Although EBV is a ubiquitous pathogen, infection only manifests as symptomatic disease in a subset of individuals who harbor the virus. The important question of whether markers of the human immune response to the virus, including antibody patterns, can identify which individuals are at increased risk of disease remains inadequately answered for most EBV-associated malignancies. Evidence does exist to indicate that altered anti-ebv antibody levels, particularly levels of immunoglobulin A (IgA) antibodies targeted against lytic and latent proteins expressed at mucosal surfaces, are predictors of the development of nasopharyngeal carcinoma (NPC) (8 11). Work is ongoing in high-incidence regions to incorporate these anti-ebv antibodies as potential screening biomarkers to aid in the detection and diagnosis of NPC (12, 13). However, the number of proteins investigated to date represents only a small percentage ( 5%) of the nearly 100 open 687

2 688 Coghill and Hildesheim reading frames in the EBV genome that could be informative, and this limited number of antigens has only been evaluated as a discriminatory tool to identify individuals who may go on to develop NPC but no other EBV-associated malignancy. Furthermore, characterizing patterns of the humoral immune response that precede disease development may shed light on the biological degree of similarity across different EBVassociated tumors, and abnormal antibody responses against the virus may reflect systemic immunodeficiency in certain cancer patients. To begin to address these issues, we conducted a review of the literature that included research studies investigating the epidemiologic association between antibodies against EBV and the risk of developing any of the following EBV-associated malignancies: NPC, gastric cancer (GC), Burkitt lymphoma (BL), non-hodgkin lymphoma (NHL), and Hodgkin lymphoma (HL). METHODS For NPC, we considered the evidence of an association between EBV serological markers (i.e., anti-ebv antibodies) measured before disease onset and the development of cancer to be strong. Therefore, an exhaustive literature review was not conducted, but key prospective studies were highlighted. In contrast, the evidence for the other 4 EBV-associated malignancies was not considered conclusive, so we conducted an exhaustive literature review using combinations of the following search terms for each malignancy: cancer x and EBV serology or EBV antibodies or EBV immune response. This keyword search was performed in PUBMED and was restricted to English-language articles published after Results generated through this keyword search are reported with an emphasis on 2 characteristics: 1) EBV proteins targeted and 2) antibody type. The EBV proteins targeted included structural proteins expressed during the lytic cycle, such as viral capsid antigen (VCA). Nonstructural proteins were also targeted, including both those crucial for viral genome maintenance inside latently infected cells, such as EBV-nuclear antigen (EBNA), and those expressed during the intracellular switch from latent phase to lytic viral replication, such as early antigen (EA) and ZEBRA. The antibody types highlighted include immunoglobulin G (IgG) and IgA. IgG antibodies are the most abundant form of antibodies produced by plasma cells in response to exposure to infectious agents and typically reflect cumulative exposure to EBV (14). IgA antibodies are produced at mucosal surfaces and are thought to be markers of more recent exposure in these compartments, such as the oral or nasopharyngeal epithelium (14). NASOPHARYNGEAL CARCINOMA EBV is a necessary cause of undifferentiated NPC, with tumors displaying an EBV latency program characterized by the expression of EBNA1 and at least 1 of the latent membrane proteins (latent membrane proteins 1 2) (15 21). In multiple case-control studies conducted in previous decades, associations between the presence of elevated anti-ebv antibody titers and NPC were found (22 27). More recent evidence from prospective studies in populations at high risk of undifferentiated NPC has confirmed these earlier observations and demonstrated that high antibody titers, particularly IgA antibodies directed against EBV structural proteins, precede the development of NPC. Taiwanese men who tested positive for VCA IgA antibodies had a higher risk of developing NPC than did men who tested negative for VCA IgA (hazard ratio = 22.0, 95% confidence interval (CI): 7.3, 66.9), even 5 or more years after antibodies were measured (hazard ratio = 13.9, 95% CI: 3.1, 61.7) (9). In a longitudinal study from China, Ji et al. (10) described a characteristic window approximately 3 years before NPC diagnosis during which VCA IgA antibodies are elevated in the majority of individuals who develop the disease. Additional recent work in a high-risk region of China has provided a strong addition to the connection between prediagnostic EBV serological markers and NPC by describing a dose-dependent relationship, linking higher VCA IgA titers to elevated risk of NPC, and reporting higher disease risk in individuals with increasing versus stable VCA IgA titers over time (hazard ratio = 12.4, 95% CI: 5.4, 28.5) (8). Furthermore, in a prospective study of members of highrisk Taiwanese multiplex families, which are defined as families with at least 2 first- or second-degree relatives who have been diagnosed with NPC, Yu et al. (11) observed elevated levels of both VCA IgA and IgA antibodies against the nonstructural EBNA1 protein before NPC diagnosis. Individuals who were positive for EBNA1 IgA at baseline experienced higher rates of NPC over follow-up (median, 6.5 years) than did persons who tested negative for EBNA1 IgA (relative risk = 6.6, 95% CI: 1.5, 61). GASTRIC CARCINOMA GC represents another EBV-associated malignancy that arises in epithelial cells, although unlike NPC, evidence of a link between antibodies against EBV and GC risk is inconsistent (Table 1, WebTable1(availableathttp://aje. oxfordjournals.org/)). One important contrast between GC and NPC is that only a minority of GC cases are associated with EBV infection, with approximately 7% 10% of tumors classified as EBV-related (28). We identified 3 prospective cohort studies (29 31) and 1 case-control study (32) that evaluated the relationship between GC and EBV serological markers, with sample sizes ranging from 54 to 185 GC cases. Two of these 4 studies did not make the important distinction between EBV-positive and EBV-negative GC but rather included a single case group. No statistically significant increases in GC risk were seen in subjects with elevated antibody titers against either structural (VCA) or nonstructural (EBNA1, EA) EBV proteins in these 2 prospective studies (29, 30). Investigators in the Chinese study (n = 185 cases) instead reported an inverse association between elevated VCA IgG antibody titers measured before diagnosis and GC risk, an association that was more pronounced after analyses were restricted to cases who were diagnosed at least 2 years after blood draw (odds ratio = 0.60, 95% CI: 0.36, 0.99) (30). This inverse association between higher VCA IgG titer and lower GC risk was 1) more pronounced for cardia GC cases than for noncardia

3 EBV Antibodies and EBV-Related Cancer Risk 689 Table 1. Gastric Carcinoma Studies That Were Reviewed, First Author, Year (Reference No.) EBV Protein Antibody Type Study Design Controls Cases IgG IgA Levine, 1995 (31) VCA IgG, IgA Prospective Positive Null Shinkura, 2000 (32) VCA IgG, IgA Case-control Positive Positive, IgA Koshiol, 2007 (30) VCA IgG, IgA Prospective Inverse Null, IgA Kim, 2009 (29) VCA IgG, IgA Prospective Null Null Abbreviations: EA, early antigen; EBNA, EBV nuclear antigen; EBV, Epstein-Barr virus; IgA, immunoglobulin A; IgG, immunoglobulin G; VCA, viral capsid antigen. a A positive association indicates that the study found a significantly higher risk of cancer associated with elevated antibody levels; a null association indicates that the study found no significant association between cancer risk and antibody levels; and an inverse association indicates that the study found a significantly lower risk of cancer associated with elevated antibody levels. cases and 2) in the opposite direction of that observed between EBV serological markers and risk for all other EBVassociated malignancies. The other 2 studies did make the important distinction between EBV-positive and EBV-negative GC tumors (31, 32). In a prospective cohort of Japanese men in whom anti-ebv antibodies were measured in serum collected an average of 13.7 years (range, years) before GC diagnosis, EBV-positive cases (n = 14) had higher geometric mean titers of IgG antibodies to both structural (VCA) and nonstructural (EBNA1) proteins than did EBV-negative cases (n = 40) and controls (n =54)(31). A second Japanese study in which antibodies were measured at the time of GC diagnosis also found higher VCA IgG geometric mean titers in EBV-positive cases (n = 64) than in EBV-negative cases (n = 59) and controls (n = 73)(32). Additionally, seropositivity rates for VCA IgA and in that case-control study were significantly higher in EBV-positive cases than in EBV-negative cases (odds ratio 7 for VCA IgA and odd ratio 20 for ). Although both studies reported increased risk of EBVpositive GC in subjects with higher VCA antibody titers, the sample size in the earlier study was limited (e.g., fewer than 20 EBV-positive cases). In addition, the data did not elucidate the importance of the timing of antibody measurement, with one study collecting blood at the time of GC diagnosis and the other measuring antibodies in blood collected years before diagnosis. Of note, results from both of these studies contrasted with the inverse association observed between VCA titers and GC risk in the larger prospective study (n = 185 cases) that did not consider EBV-positive and EBV-negative GC tumors separately (30). Unlike other EBV-associated malignancies, GC has potential precursor lesions that have been identified and can be evaluated in relation to EBV serological markers. In a longitudinal Chinese study of gastric lesions, persons with elevated baseline VCA IgG and titers had a higher likelihood of progressing to more severe gastric dysplasia 2 years later, particularly subjects with baseline diagnoses of intestinal metaplasia (33). Ascertainment of varying gastric neoplastic states of disease could allow for larger statistically powered studies to evaluate the link between EBV serological markers and gastric carcinogenesis. BURKITT LYMPHOMA BL is a lymphoid malignancy that in its endemic form is nearly uniformly associated with EBV (34) (Table2, Web Table 2) and predominantly affects children in East Africa, where a link between the presence of anti-ebv antibodies and BL risk was established through a large (n = 42,000) prospective study conducted among Ugandan children who were 0 8 years of age in the 1970s (35, 36). The original report included 14 incident cases ascertained over 5 years (35). After an additional 3 years of follow-up, 16 cases had developed (36), and titers of IgG antibodies against the structural protein VCA, although uniformly present in these children as a marker of EBV seropositivity, were significantly higher in those who developed BL. Of note, this was the only prospective study to date in which the relationship between prediagnostic EBV serological markers and risk of BL was investigated, and no studies to date have evaluated IgA antibodies. Subsequently, 1 community-based and 2 hospital-based case-control studies that evaluated IgG antibodies have been conducted in African countries, with sample sizes rangingfrom32to173blcases(37 39). These studies confirmed that children with higher VCA IgG titers had a higher BL risk regardless of the assay used, including the older immunofluorescence test (39), the more standardized enzyme-linked immunosorbent assay (38), or a multiplex bead assay (37). The most recent case-control study evaluated IgG antibody titers against nonstructural proteins such as EBNA1 and EA (37), but in accordance with the original studies conducted in Uganda (35, 36), titers were not significantly different between BL cases and controls.

4 690 Coghill and Hildesheim Table 2. Burkitt Lymphoma Studies That Were Reviewed, First Author, Year (Reference No.) EBV Protein Antibody Type Study Design Controls Cases IgG IgA Geser, 1982 (36) VCA IgG Prospective Positive NA Carpenter, 2008 (38) VCA IgG Case-control 102 b 173 Positive NA Mutalima, 2008 (39) VCA IgG Case-control 89 b 128 Positive NA Asito, 2010 (37) VCA IgG Case-control Positive NA ZEBRA IgG Abbreviations: EA, early antigen; EBNA, EBV nuclear antigen; EBV, Epstein-Barr virus; IgA, immunoglobulin A; IgG, immunoglobulin G; NA, not applicable; VCA, viral capsid antigen. a A positive association indicates that the study found a significantly higher risk of cancer associated with elevated antibody levels; NA indicates that the study did not evaluate the association. b Hospital controls. IgG antibody titers against the ZEBRA protein, which is involved in the switch from the latent phase to lytic EBV replication, were also evaluated in the most recent case-control study. No odds ratio estimates for seropositivity according to BL status were presented, but the study found that ZEBRA IgG titers were significantly elevated among the 32 BL cases compared with the 25 controls (37). Confirmation of this finding, as well as the lack of association of BL with nonstructural proteins (i.e., EBNA1), requires additional studies that include larger case numbers, are well controlled (i.e., either prospectively conducted or utilize community rather than hospital controls), and measure IgA (not just IgG) antibodies against EBV proteins. NON-HODGKIN LYMPHOMA The evidence of a link between anti-ebv antibody titers and NHL risk is inconclusive (Table 3, Web Table 3).The identification of EBV genomes and expression of oncogenic EBV proteins (i.e., latent membrane protein 1) in malignant cells has been demonstrated for only select subtypes of NHL in immunocompetent persons, including T-cell and central nervous system lymphomas (40 42). In contrast, the fraction of lymphoproliferative tumors attributable to EBV is higher for cancers that occur in patients with acquired immunodeficiency syndrome (43, 44). Nearly 100% of central nervous system lymphomas and substantive percentages (as high as 65% 75%) of certain systemic lymphomas are related to EBV infection in these immunosuppressed patients (45 47). Persistent B-cell stimulation by long-term EBV infection offers a possible mechanism via which the virus could indirectly alter the risk of B-cell lymphomas in addition to the potential directly oncogenic effects of EBV transcripts (48 51). We identified 8 studies, ranging in sample size from 11 to 491 NHL cases, that were conducted among immunocompetent individuals without stratification by tumor EBV status. Two of the prospective cohort studies investigated NHL risk in relation to IgA antibodies but found no associations (52, 53). Most studies evaluated the IgG responses to VCA and EA, and 4 of the 8 studies evaluated IgG response to EBNA in relation to NHL risk. One case-control study (54), as well as 3 of the 6 prospective studies reviewed here, reported no association between NHL risk and altered IgG antibody levels against any of these EBV proteins (52, 55, 56). Results across the case-control study and 3 prospective studies with significant findings were conflicting (53, 57, 58, 59). An analysis utilizing serum from a 4-cohort collaboration (n = 104 cases) showed a higher risk of NHL associated with elevated VCA IgG titers (hazard ratio = 3.3, 95% CI: 1.2, 8.8) and a lower risk of NHL associated with lower titers (hazard ratio = 0.2, 95% CI: 0.04, 0.9) (53). However, this risk pattern was limited to cases diagnosed within 5 years of the blood draw (n = 56), with no association observed between IgG titers and NHL diagnosed at least 5 years after the blood draw (n = 48). In contrast, a slightly smaller prospective study (n = 73 cases) found that individuals who were seropositive for were at a higher risk of developing NHL after over a decade ( 12 years) of study follow-up but observed no association with VCA IgG seropositivity (57), results that were supported by similar casecontrol findings of elevated IgG titers against EA but not VCA in blood drawn at the time of NHL diagnosis (58). Finally, the largest prospective cohort to date (n = 491 cases) reported no increase in B-cell lymphoma risk for persons with elevated IgG titers against either VCA or EA; rather, persons who were seropositive for EBNA1 IgG were observed to instead have a lower risk of NHL, irrespective of the timing between blood draw and diagnosis (59). Two of the studies reviewed here, each nested within a large US cohort, reported results by NHL subtype, and both observed suggestions of a higher risk associated with elevated anti-ebv antibodies that was specific to chronic lymphocytic leukemia, small lymphocytic lymphoma, and prolymphocytic lymphoma subtypes (55, 59). Investigators from the Physicians and Nurses Health Studies (n = 79 chronic lymphocytic leukemia/small lymphocytic lymphoma cases) reported a higher risk associated with elevated EBNA2 IgG titers (relative risk = 1.74, 95% CI: 0.99, 3.05) (55). In the Women s Health Initiative Study (n = 142 chronic lymphocytic leukemia/ small lymphocytic lymphoma /prolymphocytic lymphoma

5 EBV Antibodies and EBV-Related Cancer Risk 691 Table 3. Non-Hodgkin Lymphoma Studies That Were Reviewed, First Author, Year (Reference No.) EBV Protein Antibody Type Study Design Controls Cases IgG IgA Mueller, 1991 (53) VCA IgG, IgA Prospective Positive Null Lehitinen, 1993 (52) VCA IgG, IgA Prospective Null Null, IgA, IgA Rothman, 1997 (57) VCA IgG Prospective Positive NA Hardell, 2001 (54) VCA IgG Case-control Null NA Hardell, 2009 (58) VCA IgG Case-control Positive NA Bertrand, 2010 (55) VCA IgG Prospective Null NA Richiardi, 2010 (56) Not listed Not listed Prospective Null NA De Roos, 2013 (59) VCA IgG Prospective b Inverse NA Abbreviations: EA, early antigen; EBNA, EBV nuclear antigen; EBV, Epstein-Barr virus; IgA, immunoglobulin A; IgG, immunoglobulin G; NA, not applicable; VCA, viral capsid antigen. a A positive association indicates that the study found a significantly higher risk of cancer associated with elevated antibody levels; a null association indicates that study the found no significant association between cancer risk and antibody levels; an inverse association indicates that the study found a significantly lower risk of cancer associated with elevated antibody levels; and NA indicates that the study did not evaluate the association. b B-cell non-hodgkin lymphoma. cases), seropositivity for was significantly associated with elevated disease risk (relative risk = 1.8, 95% CI: 1.2, 2.6), although the risk was more pronounced for cases diagnosed within 4 years of blood draw (59). Additionally, work has been done to characterize patterns of multiple IgG anti-ebv antibodies simultaneously in lymphoma patients through immunoblot analysis (60). The immunoblot assay classifies the overall IgG response for a person as either 1) normal, with reactivity to a limited number of peptides (e.g., EBNA1, VCAp18, VCAp40, ZEBRA) or 2) abnormal, with recognition of a higher diversity of EBV peptides. This work has provided evidence that the abnormal humoral immune response pattern is more prevalent among persons with the chronic lymphocytic leukemia lymphoma subtype (odds ratio = 2.96, 95% CI: 2.22, 3.95). However, this immunoblot assay is not quantitative and does not provide a titer level for each individual peptide recognized. Future work is warranted to attempt to reproduce these patterned findings using a quantitative assay. HODGKIN LYMPHOMA HL is the only EBV-related malignancy other than NPC for which there is a body of evidence accumulated over time that established a strong association between altered EBV humoral immune responses and development of disease, with most work to date focused on IgG antibodies (Table 4, Web Table 4). Numerous case-control studies were conducted before 1980, many of which were summarized in an article by Evans et al. published in 1984 (61). In brief, these early studies reported significantly higher IgG titers associated with HL status, with 14 of 18 studies evaluating VCA IgG and 6 of 8 studies evaluating reporting higher prevalence of elevated titers in HL cases than in controls. Rather than reiterate what has already been summarized, below we focus on the 4 studies published on this topic since 1980, which included sample sizes ranging from 43 to 189 HL cases. Approximately 30% 40% of HL tumors in immunocompetent persons are estimated to be related to EBV (1, 62), although this proportion can be 80% or higher in persons who are infected with human immunodeficiency virus (45, 63, 64). The only study reviewed here that stratified HL tumors by EBV status reported that higher VCA, EA, and EBNA IgG antibody levels were specifically associated with elevated risk for EBV-positive HL (n = 40 cases) but not EBVnegative HL (n = 88 cases) (65). The other 3 studies (61, 66, 67) did not make a distinction between EBV-positive and EBV-negative HL tumors. However, unlike for the risks of GC and NHL, associations between EBV serological markers and HL risk were consistent even when considering one HL case group. One potential explanation is that a much higher percentage of HL tumors ( 50%) are EBV related compared with GC or NHL tumors (<10%). The VCA IgG and IgA response, as well as IgG responses to EBNA and EA, were significantly elevated in HL patients compared with controls when measured both after HL diagnosis (61) and before the development of HL (66). The prospective study reported that the association between

6 692 Coghill and Hildesheim Table 4. Hodgkin Lymphoma Studies That Were Reviewed, First Author, Year (Reference No.) EBV Protein Antibody Type Study Design Controls Cases IgG IgA Association Evans, 1984 (61) VCA IgG Case-Control 378 b 189 Positive NA Mueller, 1989 (66) VCA IgG, IgA Prospective Positive Positive Levin, 2012 (65) VCA IgG Prospective Positive NA Mueller, 2012 (67) Case-control Positive NA Abbreviations: EA, early antigen; EBNA, EBV nuclear antigen; EBV, Epstein-Barr virus; IgA, immunoglobulin A; IgG, immunoglobulin G; NA, not applicable; VCA, viral capsid antigen. a A positive association indicates that the study found a significantly higher risk of cancer associated with elevated antibody levels; NA indicates that the study did not evaluate the association. b Sibling controls. IgG antibody titers and HL risk was actually stronger after a longer latency period (66). For example, persons who developed HL were more than twice as likely to have elevated VCA IgG titers (relative risk = 2.6, 95% CI: 1.1, 6.1), but those who were diagnosed with HL at least 3 years after study enrollment were nearly 5 times as likely to have elevated VCA IgG titers (relative risk = 4.9, 95% CI: 1.5, 16.3). Interestingly, this stronger association between anti-vca antibodies measured years before HL diagnosis stands in contrast to the association of these same antibodies with short-term but not longer-term NHL risk in the very same persons (53, 66). Two studies reviewed here investigated not only single markers of the humoral immune response to EBV but also aberrant serological patterns that could reflect immune dysfunction (65, 67). The EBNA2 protein is involved in the transformation of EBV-infected lymphocytes (1, 3). High anti-ebna2 antibody titers are therefore observed after primary infection, but anti- EBNA2 antibody titers should decrease after resolution of acute infection in parallel with elevation of antibodies against the less immunogenic EBNA1 protein necessary for maintenance of latent infection (67 69). The presence of an altered ratio in which EBNA2 antibodies do not decline over the course of infection (EBNA1: EBNA2 ratio 1.0) is a phenomenon observed in patients with immunological disorders (68). In the 2 studies reviewed here, this altered EBV immune response was associated with an increased risk of HL specific to EBV-positive tumors (65) and for HL diagnosed in persons with no history of infectious mononucleosis (67). DISCUSSION The strength of evidence in the current literature for an association between anti-ebv antibody titers and risk of EBVassociated malignancies differs according to cancer diagnosis. Having elevated levels of anti-ebv IgA antibodies is related to the development of incident NPC. Although the examination of IgA antibodies is limited in relation to HL, the evidence for a link between anti-ebv IgG antibody titers and HL risk is strong. The evidence presented for elevated VCA IgG response and BL status was consistent; however, sample sizes were limited in 2 of the 4 reviewed studies, and the other 2 utilized hospital-based controls, a potentially biased comparison group. Data for GC are inconsistent, with some evidence observed for a link between the anti-vca response and EBV-positive gastric tumors. Finally, for NHL, half of the studies reviewed reported null results, although some prospective data indicate that the presence of altered anti-ebv antibodies may be important for certain subtypes of NHL. Although evidence for an association between elevated IgA antibody titers and NPC risk is strong, IgA antibodies against only 3 EBV proteins (EBNA1, VCA, and EA) have been consistently investigated in NPC studies to date (8 11). We observed that this same set of 3 proteins represented the universe of investigated EBV targets in the available literature for the other EBV-associated malignancies. This comprises only a small fraction of the approximately 100 open reading frames for EBV, meaning that little comprehensive investigation of EBV serological patterns has been conducted. The technology to explore antibody patterns and investigate responses to multiple peptides simultaneously is available (60, 70, 71). For example, one of the BL studies reviewed here utilized multiplex bead technology to quantitate the response to multiple peptides in a single assay (37). Expanding on the number of antigens evaluated could deepen our understanding of the role of the immune response to EBV in the development of these cancers. If it is consistently demonstrated that IgA antibodies are preferentially important for NPC, whereas IgG antibodies are preferentially important for HL, this might reinforce the notion that control of lytic replication at mucosal epithelial surfaces is an important predictor for NPC, whereas the expansion of the EBV-infected B-cell pool is crucial for HL. Results from lymphoma work reviewed here support the importance of investigating not just individual antibody titers but also patterns of the humoral immune response to EBV (60, 67). Elevations of specific antibodies years before disease onset may support a direct oncogenic role for EBV in certain cancers, but altered antibody patterns close to the time of diagnosis could reflect a higher prevalence of impaired immunity in persons who develop these malignancies. Both alternatives deserve further evaluation.

7 EBV Antibodies and EBV-Related Cancer Risk 693 Beyond etiological insights, the link between anti-ebv antibodies and disease risk may ultimately translate into biomarkers that can identify persons who are more likely to develop disease. Although it is premature to consider using EBV serological markers for population-level cancer prevention efforts, an endeavor that may never be justified for certain EBV-associated cancers with low incidence or small percentages of EBV-related disease, in at least one case what we have learned about etiologies of these cancers can be applied toward prevention efforts. Measurement of IgA antibody levels has been used in a large, general-population NPC screening demonstration project conducted in a high-risk region of China (72 75). Studies conducted in this region during the 1980s found rates of NPC among those that screened positive for VCA IgA that were approximately 40 times higher than those observed in the cohort at large. Unfortunately, these early efforts were not carefully controlled and do not permit accurate quantification of the potential impact of EBV serology based screening on detection or mortality rates of NPC. Since that time, enzyme-linked immunosorbent assay methods that are cheaper and easier to standardize than the immunofluorescence tests used in the 1980s have become available. Current work in this region of China is ongoing to evaluate the utility of IgA antibodies against VCA and EBNA1 as screening biomarkers, which will in turn allow us to evaluate whether implementation of EBV serology based screening can decrease the NPC mortality rate (12, 13). Our findings suggest that the relationship between the antibody response to EBV and the risk of developing EBVassociated tumors is not well characterized for at least 3 of the 5 cancers reviewed. The results from studies of the different types of cancer often conflict, evaluate antibodies against a limited number of EBV proteins, or fail to investigate multiple antibody types (e.g., IgG and IgA), thereby limiting our ability to draw definitive conclusions about the role of the humoral immune response to EBV in predicting disease risk. However, the available evidence does point to an association between EBV serological patterns and risk for certain cancers, and the availability of reproducible and affordable technology to measure antibody levels presents us with an opportunity to elucidate the role of the humoral immune response to EBV in cancer etiology and prevention moving forward. ACKNOWLEDGMENTS Author affiliation: Infections and Immunoepidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, Maryland (Anna E. Coghill, Allan Hildesheim). This research was funded by the intramural program at the National Cancer Institute. Conflict of interest: none declared. REFERENCES 1. Cohen JI. Epstein-Barr virus infection. N Engl J Med. 2000; 343(7): Hsu JL, Glaser SL. Epstein-Barr virus-associated malignancies: epidemiologic patterns and etiologic implications. Crit Rev Oncol Hematol. 2000;34(1): Young LS, Rickinson AB. Epstein-Barr virus: 40 years on. Nat Rev Cancer. 2004;4(10): Babcock GJ, Decker LL, Volk M, et al. EBV persistence in memory B cells in vivo. Immunity. 1998;9(3): Joseph AM, Babcock GJ, Thorley-Lawson DA. EBV persistence involves strict selection of latently infected B cells. J Immunol. 2000;165(6): de-thé G, Day NE, Geser A, et al. Sero-epidemiology of the Epstein-Barr virus: preliminary analysis of an international study a review. IARC Sci Publ. 1975;(11 Pt 2): Dowd JB, Palermo T, Brite J, et al. Seroprevalence of Epstein-Barr virus infection in U.S. children ages 6 19, PLoS One. 2013;8(5):e Cao SM, Liu Z, Jia WH, et al. Fluctuations of Epstein-Barr virus serological antibodies and risk for nasopharyngeal carcinoma: a prospective screening study with a 20-year follow-up. PLoS One. 2011;6(4):e Chien YC, Chen JY, Liu MY, et al. Serologic markers of Epstein-Barr virus infection and nasopharyngeal carcinoma in Taiwanese men. N Engl J Med. 2001;345(26): Ji MF, Wang DK, Yu YL, et al. Sustained elevation of Epstein-Barr virus antibody levels preceding clinical onset of nasopharyngeal carcinoma. Br J Cancer. 2007;96(4): Yu KJ, Hsu WL, Pfeiffer RM, et al. Prognostic utility of anti-ebv antibody testing for defining NPC risk among individuals from high-risk NPC families. Clin Cancer Res. 2011;17(7): Hildesheim A. Invited commentary: Epstein-Barr virusbased screening for the early detection of nasopharyngeal carcinoma: a new frontier. Am J Epidemiol. 2013;177(3): Liu Z, Ji MF, Huang QH, et al. Two Epstein-Barr virus-related serologic antibody tests in nasopharyngeal carcinoma screening: results from the initial phase of a cluster randomized controlled trial in Southern China. Am J Epidemiol. 2013; 177(3): Janeway C. Immunobiology. 7th ed. New York, NY: Garland Science, Taylor & Francis Group; Pathmanathan R, Prasad U, Sadler R, et al. Clonal proliferations of cells infected with Epstein-Barr virus in preinvasive lesions related to nasopharyngeal carcinoma. N Engl J Med. 1995; 333(11): Raab-Traub N. Epstein-Barr virus in the pathogenesis of NPC. Semin Cancer Biol. 2002;12(6): Niedobitek G, Agathanggelou A, Nicholls JM. Epstein-Barr virus infection and the pathogenesis of nasopharyngeal carcinoma: viral gene expression, tumour cell phenotype, and the role of the lymphoid stroma. Semin Cancer Biol. 1996;7(4): Vera-Sempere FJ, Burgos JS, Botella MS, et al. Immunohistochemical expression of Epstein-Barr virusencoded latent membrane protein (LMP-1) in paraffin sections of EBV-associated nasopharyngeal carcinoma in Spanish patients. Eur J Cancer B Oral Oncol. 1996;32B(3): Brooks L, Yao QY, Rickinson AB, et al. Epstein-Barr virus latent gene transcription in nasopharyngeal carcinoma cells: coexpression of EBNA1, LMP1, and LMP2 transcripts. J Virol. 1992;66(5): Fåhraeus R, Fu HL, Ernberg I, et al. Expression of Epstein-Barr virus-encoded proteins in nasopharyngeal carcinoma. Int J Cancer. 1988;42(3):

8 694 Coghill and Hildesheim 21. Pathmanathan R, Prasad U, Chandrika G, et al. Undifferentiated, nonkeratinizing, and squamous cell carcinoma of the nasopharynx. Variants of Epstein-Barr virusinfected neoplasia. Am J Pathol. 1995;146(6): de-the G, Lavoue MF, Muenz L. Differences in EBV antibody titres of patients with nasopharyngeal carcinoma originating from high, intermediate and low incidence areas. IARC Sci Publ. 1978;(20): Henle G, Henle W. Epstein-Barr virus-specific IgA serum antibodies as an outstanding feature of nasopharyngeal carcinoma. Int J Cancer. 1976;17(1): Hildesheim A, Levine PH. Etiology of nasopharyngeal carcinoma: a review. Epidemiol Rev. 1993;15(2): Lanier A, Bender T, Talbot M, et al. Nasopharyngeal carcinoma in Alaskan Eskimos Indians, and Aleuts: a review of cases and study of Epstein-Barr virus, HLA, and environmental risk factors. Cancer. 1980;46(9): Lanier AP, Bornkamm GW, Henle W, et al. Association of Epstein-Barr virus with nasopharyngeal carcinoma in Alaskan native patients: serum antibodies and tissue EBNA and DNA. Int J Cancer. 1981;28(3): Lanier AP, Henle W, Bender TR, et al. Epstein-Barr virus-specific antibody titers in seven Alaskan natives before and after diagnosis of nasopharyngeal carcinoma. Int J Cancer. 1980;26(2): Murphy G, Pfeiffer R, Camargo MC, et al. Meta-analysis shows that prevalence of Epstein-Barr virus-positive gastric cancer differs based on sex and anatomic location. Gastroenterology. 2009;137(3): Kim Y, Shin A, Gwack J, et al. Epstein-Barr virus antibody level and gastric cancer risk in Korea: a nested case-control study. Br J Cancer. 2009;101(3): Koshiol J, Qiao YL, Mark SD, et al. Epstein-Barr virus serology and gastric cancer incidence and survival. Br J Cancer. 2007; 97(11): Levine PH, Stemmermann G, Lennette ET, et al. Elevated antibody titers to Epstein-Barr virus prior to the diagnosis of Epstein-Barr-virus-associated gastric adenocarcinoma. Int J Cancer. 1995;60(5): Shinkura R, Yamamoto N, Koriyama C, et al. Epstein-Barr virus-specific antibodies in Epstein-Barr virus-positive and -negative gastric carcinoma cases in Japan. J Med Virol. 2000; 60(4): Schetter AJ, You WC, Lennette ET, et al. Association of Epstein-Barr virus antibody levels with precancerous gastric lesions in a high-risk cohort. Cancer Sci. 2008;99(2): Magrath IT. African Burkitt s lymphoma. History, biology, clinical features, and treatment. Am J Pediatr Hematol Oncol. 1991;13(2): de-thé G, Geser A, Day NE, et al. Epidemiological evidence for causal relationship between Epstein-Barr virus and Burkitt s lymphoma from Ugandan prospective study. Nature. 1978; 274(5673): Geser A, de Thé G, Lenoir G, et al. Final case reporting from the Ugandan prospective study of the relationship between EBV and Burkitt s lymphoma. Int J Cancer. 1982;29(4): Asito AS, Piriou E, Odada PS, et al. Elevated anti-zta IgG levels and EBV viral load are associated with site of tumor presentation in endemic Burkitt s lymphoma patients: a case control study. Infect Agent Cancer. 2010;5: Carpenter LM, Newton R, Casabonne D, et al. Antibodies against malaria and Epstein-Barr virus in childhood Burkitt lymphoma: a case-control study in Uganda. Int J Cancer. 2008; 122(6): Mutalima N, Molyneux E, Jaffe H, et al. Associations between Burkitt lymphoma among children in Malawi and infection with HIV, EBV and malaria: results from a case-control study. PLoS One. 2008;3(6):e Hochberg FH, Miller G, Schooley RT, et al. Centralnervous-system lymphoma related to Epstein-Barr virus. N Engl J Med. 1983;309(13): Harabuchi Y, Yamanaka N, Kataura A, et al. Epstein-Barr virus in nasal T-cell lymphomas in patients with lethal midline granuloma. Lancet. 1990;335(8682): Jones JF, Shurin S, Abramowsky C, et al. T-cell lymphomas containing Epstein-Barr viral DNA in patients with chronic Epstein-Barr virus infections. N Engl J Med. 1988;318(12): Carbone A. AIDS-related non-hodgkin s lymphomas: from pathology and molecular pathogenesis to treatment. Hum Pathol. 2002;33(4): Carbone A, Cesarman E, Spina M, et al. HIV-associated lymphomas and gamma-herpesviruses. Blood. 2009;113(6): Hamilton-Dutoit SJ, Raphael M, Audouin J, et al. In situ demonstration of Epstein-Barr virus small RNAs (EBER 1) in acquired immunodeficiency syndrome-related lymphomas: correlation with tumor morphology and primary site. Blood. 1993;82(2): MacMahon E, Glass D, Hayward D, et al. Epstein-Barr virus in AIDS-related primary central nervous system lymphoma. Lancet. 1991;338(8773): Shibata D, Weiss LM, Hernandez AM, et al. Epstein-Barr virus-associated non-hodgkin s lymphoma in patients infected with the human immunodeficiency virus. Blood. 1993;81(8): Grulich AE, Wan X, Law MG, et al. B-cell stimulation and prolonged immune deficiency are risk factors for non-hodgkin s lymphoma in people with AIDS. Aids. 2000; 14(2): Moir S, Fauci AS. Pathogenic mechanisms of B-lymphocyte dysfunction in HIV disease. J Allergy Clin Immunol. 2008; 122(1): Purdue MP, Lan Q, Bagni R, et al. Prediagnostic serum levels of cytokines and other immune markers and risk of non-hodgkin lymphoma. Cancer Res. 2011;71(14): Breen EC, Hussain SK, Magpantay L, et al. B-cell stimulatory cytokines and markers of immune activation are elevated several years prior to the diagnosis of systemic AIDS-associated non-hodgkin B-cell lymphoma. Cancer Epidemiol Biomarkers Prev. 2011;20(7): Lehtinen T, Lumio J, Dillner J, et al. Increased risk of malignant lymphoma indicated by elevated Epstein-Barr virus antibodies a prospective study. Cancer Causes Control. 1993;4(3): Mueller N, Mohar A, Evans A, et al. Epstein-Barr virus antibody patterns preceding the diagnosis of non-hodgkin s lymphoma. Int J Cancer. 1991;49(3): Hardell E, Eriksson M, Lindström G, et al. Case-control study on concentrations of organohalogen compounds and titers of antibodies to Epstein-Barr virus antigens in the etiology of non-hodgkin lymphoma. Leuk Lymphoma. 2001;42(4): Bertrand KA, Birmann BM, Chang ET, et al. A prospective study of Epstein-Barr virus antibodies and risk of non-hodgkin lymphoma. Blood. 2010;116(18): Richiardi L, De Marco L, Gillio-Tos A, et al. Persistent infection by HCV and EBV in peripheral blood mononuclear

9 EBV Antibodies and EBV-Related Cancer Risk 695 cells and risk of non-hodgkin s lymphoma. Cancer Epidemiol. 2010;34(6): Rothman N, Cantor KP, Blair A, et al. A nested case-control study of non-hodgkin lymphoma and serum organochlorine residues. Lancet. 1997;350(9073): Hardell K, Carlberg M, Hardell L, et al. Concentrations of organohalogen compounds and titres of antibodies to Epstein-Barr virus antigens and the risk for non-hodgkin lymphoma. Oncol Rep. 2009;21(6): De Roos AJ, Martinez-Maza O, Jerome KR, et al. Investigation of Epstein-Barr virus as a potential cause of B-cell non-hodgkin lymphoma in a prospective cohort. Cancer Epidemiol Biomarkers Prev. 2013;22(10): de Sanjosé S, Bosch R, Schouten T, et al. Epstein-Barr virus infection and risk of lymphoma: immunoblot analysis of antibody responses against EBV-related proteins in a large series of lymphoma subjects and matched controls. Int J Cancer. 2007;121(8): Evans AS, Gutensohn NM. A population-based case-control study of EBV and other viral antibodies among persons with Hodgkin s disease and their siblings. Int J Cancer. 1984;34(2): Glaser SL, Lin RJ, Stewart SL, et al. Epstein-Barr virusassociated Hodgkin s disease: epidemiologic characteristics in international data. Int J Cancer. 1997;70(4): Tirelli U, Errante D, Dolcetti R, et al. Hodgkin s disease and human immunodeficiency virus infection: clinicopathologic and virologic features of 114 patients from the Italian Cooperative Group on AIDS and Tumors. J Clin Oncol. 1995; 13(7): Ambinder RF. Epstein-Barr virus and Hodgkin lymphoma. Hematology Am Soc Hematol Educ Program. 2007: Levin LI, Chang ET, Ambinder RF, et al. Atypical prediagnosis Epstein-Barr virus serology restricted to EBV-positive Hodgkin lymphoma. Blood. 2012;120(18): Mueller N, Evans A, Harris NL, et al. Hodgkin s disease and Epstein-Barr virus. Altered antibody pattern before diagnosis. N Engl J Med. 1989;320(11): Mueller NE, Lennette ET, Dupnik K, et al. Antibody titers against EBNA1 and EBNA2 in relation to Hodgkin lymphoma and history of infectious mononucleosis. Int J Cancer. 2012; 130(12): Henle W, Henle G, Andersson J, et al. Antibody responses to Epstein-Barr virus-determined nuclear antigen (EBNA)-1 and EBNA-2 in acute and chronic Epstein-Barr virus infection. Proc Natl Acad Sci U S A. 1987;84(2): Rickinson ABKE. Epstein-Barr Virus. Vol 2. 5th ed. Philadelphia, PA: Lippincott Williams & Wilkins; Zheng D, Wan J, Cho YG, et al. Comparison of humoral immune responses to Epstein-Barr virus and Kaposi s sarcomaassociated herpesvirus using a viral proteome microarray. J Infect Dis. 2011;204(11): Gu AD, Lu LX, Xie YB, et al. Clinical values of multiple Epstein-Barr virus (EBV) serological biomarkers detected by xmap technology. J Transl Med. 2009;7: Zeng Y, Zhang LG, Li HY, et al. Serological mass survey for early detection of nasopharyngeal carcinoma in Wuzhou City, China. Int J Cancer. 1982;29(2): Zeng Y, Zhang LG, Wu YC, et al. Prospective studies on nasopharyngeal carcinoma in Epstein-Barr virus IgA/VCA antibody-positive persons in Wuzhou City, China. Int J Cancer. 1985;36(5): Zeng Y, Zhong JM, Li LY, et al. Follow-up studies on Epstein- Barr virus IgA/VCA antibody-positive persons in Zangwu County, China. Intervirology. 1983;20(4): Zong YS, Sham JS, Ng MH, et al. Immunoglobulin A against viral capsid antigen of Epstein-Barr virus and indirect mirror examination of the nasopharynx in the detection of asymptomatic nasopharyngeal carcinoma. Cancer. 1992;69(1):3 7.