Epstein-Barr Virus As a Marker of Survival After Hodgkin's Lymphoma: A Population-Based Study
http://www.100md.com
《临床肿瘤学》
the Northern California Cancer Center, Fremont
Stanford University Medical Center, Stanford, CA
University of North Carolina, Chapel Hill, NC
University of Pittsburgh, Pittsburgh, PA
Hartford Hospital, Hartford, CT
Johns Hopkins University, Baltimore, MD
ABSTRACT
PURPOSE: Epstein-Barr virus (EBV) in Hodgkin's lymphoma (HL) cells has been considered as a prognostic marker for this heterogeneous disease, but studies have yielded mixed findings, likely because of selected patient series and failure to acknowledge an effect of age on outcome. This study assessed survival after HL in a population-based cohort large enough to examine the joint effects of EBV with other factors including age, sex, and histologic subtype.
PATIENTS AND METHODS: Included were 922 patients with classical HL diagnosed between mid-1988 and 1997 in the Greater San Francisco Bay Area, with archived biopsy specimens assayed for EBV with immunohistochemistry and in situ hybridization. Vital status was followed through December 30, 2003 (median follow-up time, 97 months). Overall and disease-specific survival were analyzed with the Kaplan-Meier method and Cox proportional hazards regression models.
RESULTS: In children less than 15 years old, EBV presence was suggestively associated (P = .07) with favorable survival. In adults aged 15 to 44 years, EBV did not affect HL outcome, although a protective effect was suggested. In older adults (45 to 96 years), EBV presence nearly doubled the risk of overall and HL-specific mortality but only for patients with nodular sclerosis (NS) histologic subtype (hazard ratio for death = 2.5; 95% CI, 1.5 to 4.3).
CONCLUSION: In HL, EBV tumor cell presence is associated with better survival in young patients and poorer survival in older patients with NS, independent of other factors. Variation in outcome by age and histology could indicate biologically distinct disease entities. Evidence that EBV is a meaningful prognostic marker may have therapeutic relevance.
INTRODUCTION
Once a uniformly fatal disease, Hodgkin’s lymphoma (HL) now has a generally favorable outcome. However, its prognosis is well established to differ across demographic and tumor characteristics1-4; for example, the 5-year relative survival rate is 87% for patients diagnosed under age 65 years but 51% for older persons.5 Epstein-Barr virus (EBV), a ubiquitous herpes virus of known oncogenic potential detected in the malignant (Reed-Sternberg/Hodgkin’s) cells of some HL patients,6,7 also has been considered in HL outcome because characteristics of these patients (ie, young or older age at diagnosis, male sex, nonwhite race/ethnicity, and having tumors of the mixed cellularity (MC) histology subtype8-10) are correlated with poorer survival after HL. However, studies assessing the prognostic importance of EBV have yielded mixed findings.11-30 Initial research found either no impact or a suggestion of a beneficial association,11,12,14,17,24,25,30 although these investigations were based on selected patient groups, and most failed to consider modifying factors, particularly patient age at diagnosis. Our earlier population-based study with a subset of HL patients used in the present analyses was the first to report an adverse effect of EBV-positive tumor cell status on overall survival in women aged 45 years and older but little influence in younger women, after adjustment for age, stage at diagnosis, and tumor histology.13 Similar age-specific findings have recently been suggested by two small population-based studies,15,27 but one study did not adjust results for potential confounders,27 and the other did not present adjusted results.15
Determination of EBV as a prognostic marker for HL could further inform disease biology and progression and could influence treatment decisions. However, to be reliable, such a determination needs to be based on a representative patient population with uniform follow-up and to consider simultaneous influences of patient age and other prognostic indicators. Therefore, we assessed overall survival (OS) and disease-specific survival (DSS) up to 16 years after HL diagnosis in a population-based cohort large enough to examine the joint effects of EBV status, age, sex, histologic subtype, disease stage, and other factors.
PATIENTS AND METHODS
Patients
Eligible patients comprised all 1,576 patients with HL who were newly diagnosed during the period from July 1988 through December 1997 and who were residents of the nine San Francisco Bay Area counties reporting to the population-based Greater Bay Area Cancer Registry. For each HL patient, we obtained registry information routinely collected from the medical record at diagnosis on age, sex, race/ethnicity, census block group of residence, extent of disease (which permitted determination of Ann Arbor stage), presence of "B" symptoms (weight loss, night sweats, and pruritus), HIV status (available as of 1990 diagnoses), report of prior in situ or invasive cancer, and histologic subtype. In addition, we obtained registry information on treatment modality within 4 months after diagnosis (ie, any radiation, chemotherapy, or combined modality), vital status (routinely determined by the cancer registry) as of December 31, 2003, and, for the deceased, the underlying cause of death as coded by state vital statistics personnel. To enhance the accuracy of HL-related cause of death, we manually reviewed original death certificates, which were obtainable for 88% of the 255 deceased patients. HL-related deaths were defined as follows: deaths for which HL seemed to be an immediate cause (n = 7); the cause of death was lymphoma, and there was no cancer registry evidence of diagnosis with non-HL (n = 15); or HL was listed as "another significant condition contributing to death," and myelodysplasia was listed as a more proximate cause of death (n = 1). Altogether, 132 patients were determined to have an HL-related cause of death, including 23 for whom the underlying cause of death was not originally coded to HL. Because the cancer registry collects no information on patient socioeconomic status (SES), we determined the patient’s neighborhood SES using a 1990 census block group–based index, which incorporates education, income, occupation, and cost of living,31,32 that previously has demonstrated SES gradients in breast cancer1 and HL4 incidence in California and that was available for the 95% of patients whose residence at diagnosis could be geocoded. This study was approved by the Institutional Review Board of the Northern California Cancer Center.
Histologic Review and EBV Assays
For uniform diagnostic confirmation and immunohistologic classification of HL tumors and to determine their EBV status, we sought original paraffin-embedded tissue blocks for all patients and were successful for 1,178 patients (75%). Pathologist rereview (performed by R.F.D. for an earlier study of women33 and by M.L.G., F.E.C., or Bassam Smir, MD, for all other patients) was completed for the 1,110 patients with usable specimens. The 26 patients (2%) found not to have HL were excluded from analyses. Histologic classification was converted to International Classification of Diseases for Oncology, Third Edition and assigned to one of four subtypes of classical HL (nodular sclerosis [NS], lymphocyte rich [LR], MC, and lymphocyte depleted [LD]), to nodular lymphocyte predominant (LP), which is now considered a distinct B-cell neoplasm,34,35 or to HL not otherwise specified (HLNOS).
EBV status was determined by in situ hybridization for EBV-encoded RNA (EBER1) and immunohistochemistry for viral latent membrane protein (LMP1), as previously described.36 Briefly, EBER1 in situ hybridization was performed using digoxigenin-labeled EBER1 riboprobes, whereas parallel hybridization to U6 human transcripts served as a control to confirm RNA preservation in each tumor. LMP1 immunohistochemistry relied on a cocktail of antibodies (CS1-4; Dako, Carpinteria, CA).36,37 Sections of a known EBV-related HL served as a positive control in each run. Results were interpreted by experienced pathologists (R.B.M. and J.A.D. for women from the earlier study13,38 and M.L.G., F.E.C., or Bassam Smir, MD, for all other patients) working independently. A specimen was considered EBV positive if at least one definitive Reed-Sternberg/Hodgkin’s cell expressed EBER or LMP1; in most positive specimens, virtually all neoplastic cells expressed both viral gene products. Tumors were considered EBV negative if both assays were negative in Reed-Sternberg/Hodgkin’s cells or if the single successful assay was negative. A result was determined for tumors of 1,038 patients with definitive HL.
Final Study Population
After excluding 63 patients with HIV-related HL39 because of its extremely poor survival and strong association with EBV,40 one patient with missing age at diagnosis, and two patients with an unknown/zero value of survival time, we analyzed survival for 922 patients who were observed for 0.1 to 195 months after diagnosis (median, 97 months). Fifty additional patients with nodular LP HL were evaluated separately.
Statistical Analyses
Analyses assessed OS, which evaluated the risk of death from all causes, and DSS, which evaluated the risk of death from HL. For deceased patients, survival time was measured in months from the date of diagnosis to the date of death of any cause for OS and to the date of death from HL for DSS (patients dying from other causes were censored at the time of death). Patients alive at the study end date (December 31, 2003) were censored at this time or at date of last follow-up (ie, last known contact). Seventy percent of patients had a follow-up date within 1 year of the study end date, whereas 80% had a follow-up within 2 years. EBV status did not differ for patients with and without follow-up information within 2 years.
Differences in OS and DSS Kaplan-Meier survival curves between EBV-positive and EBV-negative patients were compared statistically using the log-rank test. Kaplan-Meier curves were stratified by 15-year age groups (15 to 29, 30 to 44, 45 to 59, 60 to 74, and 75+ years), sex, race/ethnicity (non-Hispanic white v other), histologic subtype, presence of B symptoms, stage of disease, and initial treatment. Given our previously noted deficit in OS for EBV-positive HL only among women more than 45 years old1 and other survival differences by age,1 we ultimately conducted survival analyses separately for children (0 to 14 years old), younger adults (15 to 44 years old), and older adults (45 years old and older). Because of small numbers, lymphocyte-rich and lymphocyte-depleted histologic subtypes were combined for multivariate analyses.
To evaluate the impact of EBV status on survival controlling for other prognostic factors, we used Cox proportional hazards regression to determine hazard ratios (HR), which estimate the instantaneous relative risk of death averaged over the entire time period, and associated 95% CIs. Multivariate analyses included variables significant at P < .15 in univariate models or with a priori hypotheses for inclusion (eg, age, race/ethnicity, and sex). In all models, the proportional hazards assumption was assessed by visual inspection of the survival curves (log [–log] of the survival distribution function by log [months]). Effect modification between covariates was examined by entering interaction terms into the multivariate model. All HRs presented were adjusted for age and race/ethnicity unless otherwise noted. All analyses were conducted using SAS version 9 software (SAS Institute Inc, Cary, NC).
RESULTS
Cohort Characteristics
Of the 922 patients, 67% were between the ages of 15 and 44 years at diagnosis, and 73% were of non-Hispanic white race/ethnicity. The group included nearly equal proportions of males and females. Seventy-four percent of the patients had the NS histologic subtype, and 61% had Ann Arbor stage I or II disease at diagnosis. As of the study end date, 74% of the cohort was alive.
Table 1 demonstrates that the proportion of all patients with EBV-positive tumors was 27%, but this rate varied by age (40% in children and older adults and 19% in young adults). Overall, EBV-positive patients were more likely than EBV-negative patients to be nonwhite and male, live in lower SES neighborhoods, and have the MC histologic subtype. In children, EBV-positive patients were less likely than EBV-negative patients to have stage IV disease or B symptoms. In younger adults, EBV-positive patients were somewhat more likely to have a later stage of disease (stage III/IV) at diagnosis. In older adults, EBV-positive patients were more likely to have B symptoms and somewhat more likely to have had a prior cancer than EBV-negative patients. The type of initial treatment did not differ significantly between EBV-positive and EBV-negative patients within groups defined by age (age 15 to 44 years or 45+ years) or stage at diagnosis (stage I/II or III/IV; data not shown). Vital status also differed by age and EBV status in children (deaths in 6% of EBV-positive and 30% of EBV-negative patients) and older adults (deaths in 74% of EBV-positive and 53% of EBV-negative patients) but not in young adults (deaths in 12% of both EBV-positive and EBV-negative patients).
Survival Characteristics
The relationship of EBV to both OS and DSS differed markedly across age group. For children, EBV-positive HL seemed to have better OS and DSS (all deaths were a result of HL) than EBV-negative HL, as shown in Figure 1 (log-rank test, P = .07) and by Cox regression analysis (HR = 0.18; 95% CI, 0.02 to 1.7). However, these findings were not statistically significant, and the small numbers of deaths in this age group precluded further multivariate analyses.
In young adults, EBV status was not significantly associated with OS (Fig 2A; log-rank test, P = .76) or DSS (Fig 2B; log-rank test, P = .40). Cox models confirmed that EBV-positive tumor status did not impact HL outcome for OS overall or in data stratified by the NS or MC histologic subtypes (Table 2) or for DSS (HR = 0.59; 95% CI, 0.25 to 1.4). EBV presence also did not significantly influence survival in patient subgroups defined by race/ethnicity, sex, stage at diagnosis, presence of B symptoms, initial treatment, or age (for 346 patients 15 to 29 years old: HR for OS = 1.3; 95% CI, 0.62 to 2.9; HR for DSS = 0.73; 95% CI, 0.21 to 2.5; for 271 patients 30 to 44 years old: HR for OS = 0.37; 95% CI, 0.13 to 1.1; HR for DSS = 0.40; 95% CI, 0.12 to 1.4), although data were suggestive of better OS for 30 to 44 year olds.
In older adults, OS was substantially poorer for patients with EBV-positive than EBV-negative tumors (Fig 3A; log-rank test, P < .01). This deficit persisted irrespective of sex (men: HR = 1.4; 95% CI, 0.93 to 2.3; women: HR = 1.6; 95% CI, 1.0 to 2.7), stage at diagnosis, initial treatment, sex, race/ethnicity, and presence of B symptoms (data not shown). It also persisted across smaller age groups, albeit less so for the oldest age group (for 111 patients 45 to 59 years old: HR = 1.9; 95% CI, 1.0 to 3.7; for 105 patients 60 to 74 years old: HR = 1.6; 95% CI, 0.97 to 2.5; for 53 patients 75 to 96 years old: HR = 1.3; 95% CI, 0.70 to 2.4). Results were similar for DSS (Fig 3B) because most of the survival differences by EBV status were a result of HL-related deaths. In data stratified by histologic subtype, the deficit in OS was apparent for patients with NS (Fig 4A; log-rank test, P < .01) but not for patients with MC (Fig 4B; log-rank test, P = .80). Survival was better for patients with EBV-negative NS (76% at 50 months) than for patients with EBV-positive NS, EBV-positive MC, or EBV-negative MC, in whom the survival rate was lower and similar (approximately 50% at 50 months). Cox models (Table 3) confirmed that the hazard of death from any cause was elevated in EBV-positive compared with EBV-negative older adults, reflecting a 2.5-fold increase in NS patients. Similar results occurred for DSS, with HRs (adjusted for age, race/ethnicity, sex, histologic subtype, B symptoms, and disease stage) of 1.8 (95% CI, 1.1 to 2.9) overall, 2.4 (95% CI, 1.2 to 5.1) for NS, and 1.5 (95% CI, 0.6 to 3.7) for MC.
Among patients with nodular LP HL, 66% were male, 60% were white, and 10% were EBV positive. Over the follow-up period, 22% died; OS was suggestively poorer for EBV-positive than EBV-negative patients (log-rank test, P = .09; HR = 2.0; 95% CI, 0.49 to 8.5).
DISCUSSION
This population-based study has confirmed that EBV presence in HL tumor cells is a meaningful molecular marker of outcome after HL in both males and females, independent of other factors, but that its influence varies by patient age at diagnosis and HL histologic subtype. In children, EBV-positive status was suggestively associated with more favorable survival than EBV-negative status. In young adults, EBV did not significantly affect HL outcome, although the data suggested a possible protective effect in patients aged 30 to 44 years at diagnosis. In contrast, in older adults, EBV presence nearly doubled the risk of overall and HL-specific mortality, even after adjustment for age, stage at diagnosis, and other factors, but this effect was limited to patients with NS histologic subtype. Overall, our findings suggest that tumor cell EBV plays a complex role in the progression of HL but might serve as a useful indicator of prognosis in some groups, including older adults with NS and, possibly, children.
The relationship of EBV presence in HL tumor cells to survival after HL has been evaluated in 19 published studies, with varied results (Table 4). 11-30 The range of findings is likely attributable in part to differences in patient populations across studies, infrequent consideration of age, limited statistical power as a result of small sample sizes, and differing outcome measures (OS, DSS, or disease/failure-free survival). However, our current findings are consistent with those of studies using unselected patient groups and acknowledging age effects. The two prior studies looking at children reported a favorable effect of EBV-positive HL on survival,16,26 as we have noted here. Prior studies composed primarily of young adults found no EBV-related difference in survival,12,14,21,41 whereas three studies reported favorable OS among patients with EBV-positive tumors.23,24,26 Flavell et al18 found a trend towards longer disease-free survival in patients 15 to 34 years of age, and Glavina-Durdov et al19 observed a longer disease-free survival in EBV-positive patients under 35 years of age with an opposite finding in older adults. These findings are generally consistent with our null associations in young adults and suggestive observations for persons aged 30 to 44 years.18,19 Two prior population-based studies of adults greater than age 60 years found poorer OS15,27 and DSS15 in EBV-positive compared with EBV-negative patients, which is similar to our finding in adults older than 45 years.
However, few studies have considered the modifying effect of histologic subtype for EBV presence on HL outcome. Five studies reported no differences across subtypes but did not present their findings in detail,11,12,17,29,30 and one study combined NS and MC subtypes.19 No prior study presented data stratified jointly by age and histologic subtype. We have shown a significant interaction of age, EBV status, and histologic subtype with survival after HL, with poorer outcome for adults diagnosed at greater than age 44 years with EBV-positive NS, which is a group that comprises one third of all older HL patients.42 The presence of EBV in HL has been one of the criteria used for therapies such as EBV-specific cell infusion.43 Our findings suggest that detection of EBV in tumor cells might be relevant to guiding additional therapies, perhaps including high-dose therapy, in the older age group in whom the presence of EBV confers a poor prognosis. However, comorbidities and poor tolerance to cytotoxic therapies may limit the applicability of cytotoxic intensification in this age group.
EBV could promote tumor maintenance through a variety of mechanisms, including the same pathways it uses to persist in healthy hosts, namely increasing host cell proliferation, decreasing apoptosis, and eluding the immune system in part by downregulating HLA-mediated presentation of viral peptides. The variation in impact of tumor cell EBV on HL outcomes in patient groups defined by age and histology could indicate that these groups represent biologically distinct disease entities, as suggested by ample epidemiologic evidence6 and differences in EBV-specific cellular immunity in young and older adults.44 The poorer general immunocompetence of older patients may facilitate faster growth of an EBV-positive tumor than what might occur in children or young adults; in younger patients, EBV antigens might stimulate more robust anti-EBV or antitumor immune responses.45,46 Viral gene expression might be expected to be immunogenic at all ages because EBV-specific cytotoxic T cells can be isolated from most HL patients.47 However, EBV-positive HL tumor cells may escape immunosurveillance.48 LMP1 expression may alter immune response insofar as the viral protein induces expression of interleukin (IL) -10 or IL-6, which are cytokines that perturb the cellular immune response.49,50 Thus, high concentrations of IL-10 within the tumor mass may suppress EBV-specific cytotoxicity sufficiently to allow malignant cells to thrive, particularly given age-related declines in immunocompetence. Alternatively, cytokine imbalances could have other detrimental influences on HL survival.51 These mechanisms could be explored with more basic research addressing EBV-specific immune responses and the ways they are modified by shifting T-helper cytokine balance or other aspects of immunocompetence that might differ between young and older adult HL patients.
This analysis had the advantages of using a large group of HL patients of all ages identified from a population-based cancer registry with uniformly collected demographic and clinical data, expert histopathologic rereview, and relatively long follow-up. However, because we could not obtain tumor blocks for all eligible patients, patients in our analysis differed from those who were excluded. Those patients who were included were more likely to be female because the earlier study of women52 occurred when specimen retrieval was easier. Among children, included patients, compared with children who were not included, were more likely to have HL not otherwise specified (3% v 14%, respectively) and higher SES (38% v 26% in the highest quintile, respectively). Among older adults, included versus nonincluded patients were more like to be non-Hispanic white (75% v 65%, respectively) and to have characteristics associated with better survival (NS histologic subtype, earlier stage disease, and no B symptoms), in part, because we lacked patient consent for block release from deceased patients. These differences are likely to have affected our statistical power by reducing our sample size and may have biased our findings toward the null, especially in older adults in whom differences were most marked. Our study lacked information on disease recurrence, detailed treatment, and comorbid conditions, which are not collected by the cancer registry; however, EBV tumor cell status has been shown to have similar associations with disease-free and failure-free survival and with OS and/or DSS in most14,15,23,25,26 but not all18,19,21 studies with more detailed clinical information. Moreover, we found no association between first course of treatment and EBV presence when considering age at diagnosis and stage of disease, which is similar to a Swedish study of uniformly treated patients.15
Our findings that EBV in HL tumors decreased OS and DSS in older NS patients while conferring a survival advantage in children and possibly young adults indicate that EBV histochemical stains performed in these patient groups may assist in prognosis and clinical management decisions. The findings also support further investigation of EBV as a prognostic marker for HL. Understanding the mechanisms behind these associations will not only improve our understanding of the complexities of HL pathogenesis but may also be helpful in developing new therapeutic strategies targeting EBV directly or in designing better conventional regimens. However, future studies should include larger numbers of children and young adults to study the joint effects of age and EBV status with other prognostic factors and should include detailed clinical information to identify subsets of patients who respond differentially to therapy and, thus, to enhance the therapeutic applicability of these observations.
Authors' Disclosures of Potential Conflicts of Interest
Margaret L. GulleyUS Public Health Funded Project (NIH) (B)Richard F. AmbinderNCI (C)
NOTES
Supported by National Cancer Institute (NCI) Grants No. NCI R29 CA-50381 (S.L.G.), NCI RO3 CA-63245 (S.L.G., R.F.A.), and N01-CN-65107 (S.L.G.). The collection of Cancer incidence data used in this study was supported by the California Department of Health Services as part of the statewide cancer reporting program mandated by California Health and Safety Code Section 103885; the National Cancer Institute's Surveillance, Epidemiology and End Results Program under contract N01-PC-35136 awarded to the Northern California Cancer Center, contract N01-PC-35139 awarded to the University of Southern California, and contract N02-PC-15105 awarded to the Public Health Institute; and the Centers for Disease Control and Prevention's Nationals Program of Cancer Registries, under agreement #U55/CCR921930-02 awarded to the Public Health Institute. The ideas and opinions expressed herein are those of the authors, and endorsement by the state of California, Department of Health Services, the National Cancer Institute, and the Centers for Disease Control and Prevention or their contractors and subcontractors is not intended, nor should it be inferred.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Clarke CA, Glaser SL, Prehn AW: Age-specific survival after Hodgkin's disease in a population-based cohort (United States). Cancer Causes Control 12:803-812, 2001
Kennedy BJ, Loeb V Jr, Peterson V, et al: Survival in Hodgkin's disease by stage and age. Med Pediatr Oncol 20:100-104, 1992
Kennedy BJ, Fremgen AM, Menck HR: Hodgkin's disease survival by stage and age. J Am Geriatr Soc 48:315-317, 2000
Walker A, Schoenfeld ER, Lowman JT, et al: Survival of the older patient compared with the younger patient with Hodgkin's disease: Influence of histologic type, staging, and treatment. Cancer 65:1635-1640, 1990
Ries LAG, Elsner MP, Kosary CL, et al: (eds): SEER Cancer Statistics Review, 1973-1999. Bethesda, MD, National Cancer Institute, 2002
Mueller N: Hodgkin's disease, in Schottenfeld D, Fraumeni JF Jr (eds): Cancer Epidemiology and Prevention (ed 2). New York, NY, Oxford University Press, 1996, pp 893-919
Mueller N, Grufferman S: The epidemiology of Hodgkin's disease, in Mauch PM, Armitage JO, Diehl V, et al (eds): Hodgkin's Disease. Philadelphia, Lippincott Williams & Wilkins, 1999, pp 61-77
Flavell KJ, Biddulph JP, Constandinou CM, et al: Variation in the frequency of Epstein-Barr virus-associated Hodgkin's disease with age. Leukemia 14:748-753, 2000
Glaser SL, Jarrett RF: The epidemiology of Hodgkin's disease. Baillieres Clin Haematol 9:401-416, 1996
Jarrett RF: Risk factors for Hodgkin's lymphoma by EBV status and significance of detection of EBV genomes in serum of patients with EBV-associated Hodgkin's lymphoma. Leuk Lymphoma 44:S27-S32, 2003 (suppl 3)
Armstrong AA, Lennard A, Alexander FE, et al: Prognostic significance of Epstein-Barr virus association in Hodgkin's disease. Eur J Cancer 30A:1045-1046, 1994
Axdorph U, Porwit-MacDonald A, Sjoberg J, et al: Epstein-Barr virus expression in Hodgkin's disease in relation to patient characteristics, serum factors and blood lymphocyte function. Br J Cancer 81:1182-1187, 1999
Clarke CA, Glaser SL, Dorfman RF, et al: Epstein-Barr virus and survival after Hodgkin's disease in a population-based series of women. Cancer 91:1579-1587, 2001
Enblad G, Sandvej K, Lennette E, et al: Lack of correlation between EBV serology and presence of EBV in the Hodgkin and Reed-Sternberg cells of patients with Hodgkin's disease. Int J Cancer 72:394-397, 1997
Enblad G, Sandvej K, Sundstrom C, et al: Epstein-Barr virus distribution in Hodgkin's disease in an unselected Swedish population. Acta Oncol 38:425-429, 1999
Engel M, Essop MF, Close P, et al: Improved prognosis of Epstein-Barr virus associated childhood Hodgkin's lymphoma: Study of 47 South African cases. J Clin Pathol 53:182-186, 2000
Fellbaum C, Hansmann ML, Niedermeyer H, et al: Influence of Epstein-Barr virus genomes on patient survival in Hodgkin's disease. Am J Clin Pathol 98:319-323, 1992
Flavell KJ, Billingham LJ, Biddulph JP, et al: The effect of Epstein-Barr virus status on outcome in age- and sex-defined subgroups of patients with advanced Hodgkin's disease. Ann Oncol 14:282-290, 2003
Glavina-Durdov M, Jakic-Razumovic J, Capkun V, et al: Assessment of the prognostic impact of the Epstein-Barr virus-encoded latent membrane protein-1 expression in Hodgkin's disease. Br J Cancer 84:1227-1234, 2001
Herling M, Rassidakis GZ, Medeiros LJ, et al: Expression of Epstein-Barr virus latent membrane protein-1 in Hodgkin and Reed-Sternberg cells of classical Hodgkin's lymphoma: Associations with presenting features, serum interleukin 10 levels, and clinical outcome. Clin Cancer Res 9:2114-2120, 2003
Krugmann J, Tzankov A, Gschwendtner A, et al: Longer failure-free survival interval of Epstein-Barr virus-associated classical Hodgkin's lymphoma: A single-institution study. Mod Pathol 16:566-573, 2003
Levy A, Diomin V, Gopas J, et al: Hodgkin's lymphoma in the Bedouin of southern Israel: Epidemiological and clinical features. Isr Med Assoc J 2:501-503, 2000
Montalban C, Abraira V, Morente M, et al: Epstein-Barr virus-latent membrane protein 1 expression has a favorable influence in the outcome of patients with Hodgkin's disease treated with chemotherapy. Leuk Lymphoma 39:563-572, 2000
Morente MM, Piris MA, Abraira V, et al: Adverse clinical outcome in Hodgkin's disease is associated with loss of retinoblastoma protein expression, high Ki67 proliferation index, and absence of Epstein-Barr virus-latent membrane protein 1 expression. Blood 90:2429-2936, 1997
Murray PG, Billingham LJ, Hassan HT, et al: Effect of Epstein-Barr virus infection on response to chemotherapy and survival in Hodgkin's disease. Blood 94:442-447, 1999
Naresh KN, Johnson J, Srinivas V, et al: Epstein-Barr virus association in classical Hodgkin's disease provides survival advantage to patients and correlates with higher expression of proliferation markers in Reed-Sternberg cells. Ann Oncol 11:91-96, 2000
Stark GL, Wood KM, Jack F, et al: Hodgkin's disease in the elderly: A population-based study. Br J Haematol 119:432-440, 2002
Vassallo J, Metze K, Traina F, et al: Expression of Epstein-Barr virus in classical Hodgkin's lymphomas in Brazilian adult patients. Haematologica 86:1227-1228, 2001
Vassallo J, Metze K, Traina F, et al: The prognostic relevance of apoptosis-related proteins in classical Hodgkin's lymphomas. Leuk Lymphoma 44:483-488, 2003
Vestlev PM, Pallesen G, Sandvej K, et al: Prognosis of Hodgkin's disease is not influenced by Epstein-Barr virus latent membrane protein. Int J Cancer 50:670-671, 1992
Clarke CA, Glaser SL, Keegan TH, et al: Neighborhood socioeconomic status and Hodgkin's lymphoma incidence in California. Cancer Epidemiol Biomarkers Prev 14:1441-1447, 2005
Yost K, Perkins C, Cohen R, et al: Socioeconomic status and breast cancer incidence in California for different race/ethnic groups. Cancer Causes Control 12:703-711, 2001
Glaser SL, Dorfman RF, Clarke CA: Expert review of the diagnosis and histologic classification of Hodgkin's disease in a population-based cancer registry: Interobserver reliability and impact on incidence and survival rates. Cancer 92:218-224, 2001
Diehl V, Sextro M, Franklin J, et al: Clinical presentation, course, and prognostic factors in lymphocyte-predominant Hodgkin's disease and lymphocyte-rich classical Hodgkin's disease: Report from the European Task Force on Lymphoma Project on Lymphocyte-Predominant Hodgkin's Disease. J Clin Oncol 17:776-783, 1999
Pileri SA, Ascani S, Leoncini L, et al: Hodgkin's lymphoma: The pathologist's viewpoint. J Clin Pathol 55:162-176, 2002
Gulley ML, Glaser SL, Craig FE, et al: Guidelines for interpreting EBER in situ hybridization and LMP1 immunohistochemical tests for detecting Epstein-Barr virus in Hodgkin's disease. Am J Clin Pathol 117:259-267, 2002
Murray PG, Constandinou CM, Crocker J, et al: Analysis of major histocompatibility complex class I, TAP expression, and LMP2 epitope sequence in Epstein-Barr virus-positive Hodgkin's disease. Blood 92:2477-2483, 1998
Glaser SL, Keegan THM, Clarke CA, et al: Exposure to childhood infections and risk of EBV-defined Hodgkin's lymphoma in women. Int J Cancer 115:599-605, 2005
Clarke C, Glaser SL: Population-based surveillance of HIV-associated cancers: Utility of cancer registry data. J Acquir Immune Defic Syndr 36:1083-1091, 2004
Glaser SL, Clarke CA, Gulley ML, et al: Population-based patterns of human immunodeficiency virus-related Hodgkin lymphoma in the Greater San Francisco Bay Area, 1988-1998. Cancer 98:300-309, 2003
Herrmann K, Niedobitek G: Lack of evidence for an association of Epstein-Barr virus infection with breast carcinoma. Breast Cancer Res 5:R13-R17, 2003
Surveillance Epidemiology and End Results (SEER) Program: SEER*Stat Database: Incidence—SEER 9 Regs Public-Use, Nov 2003 Sub (1973-2001). Bethesda, MD, National Cancer Institute, Division of Cancer Control and Population Sciences, Surveillance Research Program, Cancer Statistics Branch, November, 2003
Rooney CM, Bollard C, Huls MH, et al: Immunotherapy for Hodgkin's disease. Ann Hematol 81:S39-S42, 2002 (suppl 2)
Gandhi MK, Tellam JT, Khanna R: Epstein-Barr virus-associated Hodgkin's lymphoma. Br J Haematol 125:267-281, 2004
Armstrong AA, Alexander FE, Cartwright R, et al: Epstein-Barr virus and Hodgkin's disease: Further evidence for the three disease hypothesis. Leukemia 12:1272-1276, 1998
Frisan T, Sjoberg J, Dolcetti R, et al: Local suppression of Epstein-Barr virus (EBV)-specific cytotoxicity in biopsies of EBV-positive Hodgkin's disease. Blood 86:1493-1501, 1995
Roskrow MA, Suzuki N, Gan Y, et al: Epstein-Barr virus (EBV)-specific cytotoxic T lymphocytes for the treatment of patients with EBV-positive relapsed Hodgkin's disease. Blood 91:2925-2934, 1998
Rickinson AB, Kieff E: Epstein-Barr virus, in Fields BN, Knipe DM, Howley PM, et al (eds): Field's Virology (ed 3). Philadelphia, PA, Lippincott-Raven Publishers, 1996, pp 2397-2446
Herbst H, Foss HD, Samol J, et al: Frequent expression of interleukin-10 by Epstein-Barr virus-harboring tumor cells of Hodgkin's disease. Blood 87:2918-2929, 1996
Herbst H, Samol J, Foss HD, et al: Modulation of interleukin-6 expression in Hodgkin and Reed-Sternberg cells by Epstein-Barr virus. J Pathol 182:299-306, 1997
Seymour JF, Talpaz M, Hagemeister FB, et al: Clinical correlates of elevated serum levels of interleukin 6 in patients with untreated Hodgkin's disease. Am J Med 102:21-28, 1997
Glaser SL, Clarke CA, Stearns CB, et al: Age variation in Hodgkin's disease risk factors in older women: Evidence from a population-based case-control study. Leuk Lymphoma 42:997-1004, 2001(Theresa H.M. Keegan, Sall)
Stanford University Medical Center, Stanford, CA
University of North Carolina, Chapel Hill, NC
University of Pittsburgh, Pittsburgh, PA
Hartford Hospital, Hartford, CT
Johns Hopkins University, Baltimore, MD
ABSTRACT
PURPOSE: Epstein-Barr virus (EBV) in Hodgkin's lymphoma (HL) cells has been considered as a prognostic marker for this heterogeneous disease, but studies have yielded mixed findings, likely because of selected patient series and failure to acknowledge an effect of age on outcome. This study assessed survival after HL in a population-based cohort large enough to examine the joint effects of EBV with other factors including age, sex, and histologic subtype.
PATIENTS AND METHODS: Included were 922 patients with classical HL diagnosed between mid-1988 and 1997 in the Greater San Francisco Bay Area, with archived biopsy specimens assayed for EBV with immunohistochemistry and in situ hybridization. Vital status was followed through December 30, 2003 (median follow-up time, 97 months). Overall and disease-specific survival were analyzed with the Kaplan-Meier method and Cox proportional hazards regression models.
RESULTS: In children less than 15 years old, EBV presence was suggestively associated (P = .07) with favorable survival. In adults aged 15 to 44 years, EBV did not affect HL outcome, although a protective effect was suggested. In older adults (45 to 96 years), EBV presence nearly doubled the risk of overall and HL-specific mortality but only for patients with nodular sclerosis (NS) histologic subtype (hazard ratio for death = 2.5; 95% CI, 1.5 to 4.3).
CONCLUSION: In HL, EBV tumor cell presence is associated with better survival in young patients and poorer survival in older patients with NS, independent of other factors. Variation in outcome by age and histology could indicate biologically distinct disease entities. Evidence that EBV is a meaningful prognostic marker may have therapeutic relevance.
INTRODUCTION
Once a uniformly fatal disease, Hodgkin’s lymphoma (HL) now has a generally favorable outcome. However, its prognosis is well established to differ across demographic and tumor characteristics1-4; for example, the 5-year relative survival rate is 87% for patients diagnosed under age 65 years but 51% for older persons.5 Epstein-Barr virus (EBV), a ubiquitous herpes virus of known oncogenic potential detected in the malignant (Reed-Sternberg/Hodgkin’s) cells of some HL patients,6,7 also has been considered in HL outcome because characteristics of these patients (ie, young or older age at diagnosis, male sex, nonwhite race/ethnicity, and having tumors of the mixed cellularity (MC) histology subtype8-10) are correlated with poorer survival after HL. However, studies assessing the prognostic importance of EBV have yielded mixed findings.11-30 Initial research found either no impact or a suggestion of a beneficial association,11,12,14,17,24,25,30 although these investigations were based on selected patient groups, and most failed to consider modifying factors, particularly patient age at diagnosis. Our earlier population-based study with a subset of HL patients used in the present analyses was the first to report an adverse effect of EBV-positive tumor cell status on overall survival in women aged 45 years and older but little influence in younger women, after adjustment for age, stage at diagnosis, and tumor histology.13 Similar age-specific findings have recently been suggested by two small population-based studies,15,27 but one study did not adjust results for potential confounders,27 and the other did not present adjusted results.15
Determination of EBV as a prognostic marker for HL could further inform disease biology and progression and could influence treatment decisions. However, to be reliable, such a determination needs to be based on a representative patient population with uniform follow-up and to consider simultaneous influences of patient age and other prognostic indicators. Therefore, we assessed overall survival (OS) and disease-specific survival (DSS) up to 16 years after HL diagnosis in a population-based cohort large enough to examine the joint effects of EBV status, age, sex, histologic subtype, disease stage, and other factors.
PATIENTS AND METHODS
Patients
Eligible patients comprised all 1,576 patients with HL who were newly diagnosed during the period from July 1988 through December 1997 and who were residents of the nine San Francisco Bay Area counties reporting to the population-based Greater Bay Area Cancer Registry. For each HL patient, we obtained registry information routinely collected from the medical record at diagnosis on age, sex, race/ethnicity, census block group of residence, extent of disease (which permitted determination of Ann Arbor stage), presence of "B" symptoms (weight loss, night sweats, and pruritus), HIV status (available as of 1990 diagnoses), report of prior in situ or invasive cancer, and histologic subtype. In addition, we obtained registry information on treatment modality within 4 months after diagnosis (ie, any radiation, chemotherapy, or combined modality), vital status (routinely determined by the cancer registry) as of December 31, 2003, and, for the deceased, the underlying cause of death as coded by state vital statistics personnel. To enhance the accuracy of HL-related cause of death, we manually reviewed original death certificates, which were obtainable for 88% of the 255 deceased patients. HL-related deaths were defined as follows: deaths for which HL seemed to be an immediate cause (n = 7); the cause of death was lymphoma, and there was no cancer registry evidence of diagnosis with non-HL (n = 15); or HL was listed as "another significant condition contributing to death," and myelodysplasia was listed as a more proximate cause of death (n = 1). Altogether, 132 patients were determined to have an HL-related cause of death, including 23 for whom the underlying cause of death was not originally coded to HL. Because the cancer registry collects no information on patient socioeconomic status (SES), we determined the patient’s neighborhood SES using a 1990 census block group–based index, which incorporates education, income, occupation, and cost of living,31,32 that previously has demonstrated SES gradients in breast cancer1 and HL4 incidence in California and that was available for the 95% of patients whose residence at diagnosis could be geocoded. This study was approved by the Institutional Review Board of the Northern California Cancer Center.
Histologic Review and EBV Assays
For uniform diagnostic confirmation and immunohistologic classification of HL tumors and to determine their EBV status, we sought original paraffin-embedded tissue blocks for all patients and were successful for 1,178 patients (75%). Pathologist rereview (performed by R.F.D. for an earlier study of women33 and by M.L.G., F.E.C., or Bassam Smir, MD, for all other patients) was completed for the 1,110 patients with usable specimens. The 26 patients (2%) found not to have HL were excluded from analyses. Histologic classification was converted to International Classification of Diseases for Oncology, Third Edition and assigned to one of four subtypes of classical HL (nodular sclerosis [NS], lymphocyte rich [LR], MC, and lymphocyte depleted [LD]), to nodular lymphocyte predominant (LP), which is now considered a distinct B-cell neoplasm,34,35 or to HL not otherwise specified (HLNOS).
EBV status was determined by in situ hybridization for EBV-encoded RNA (EBER1) and immunohistochemistry for viral latent membrane protein (LMP1), as previously described.36 Briefly, EBER1 in situ hybridization was performed using digoxigenin-labeled EBER1 riboprobes, whereas parallel hybridization to U6 human transcripts served as a control to confirm RNA preservation in each tumor. LMP1 immunohistochemistry relied on a cocktail of antibodies (CS1-4; Dako, Carpinteria, CA).36,37 Sections of a known EBV-related HL served as a positive control in each run. Results were interpreted by experienced pathologists (R.B.M. and J.A.D. for women from the earlier study13,38 and M.L.G., F.E.C., or Bassam Smir, MD, for all other patients) working independently. A specimen was considered EBV positive if at least one definitive Reed-Sternberg/Hodgkin’s cell expressed EBER or LMP1; in most positive specimens, virtually all neoplastic cells expressed both viral gene products. Tumors were considered EBV negative if both assays were negative in Reed-Sternberg/Hodgkin’s cells or if the single successful assay was negative. A result was determined for tumors of 1,038 patients with definitive HL.
Final Study Population
After excluding 63 patients with HIV-related HL39 because of its extremely poor survival and strong association with EBV,40 one patient with missing age at diagnosis, and two patients with an unknown/zero value of survival time, we analyzed survival for 922 patients who were observed for 0.1 to 195 months after diagnosis (median, 97 months). Fifty additional patients with nodular LP HL were evaluated separately.
Statistical Analyses
Analyses assessed OS, which evaluated the risk of death from all causes, and DSS, which evaluated the risk of death from HL. For deceased patients, survival time was measured in months from the date of diagnosis to the date of death of any cause for OS and to the date of death from HL for DSS (patients dying from other causes were censored at the time of death). Patients alive at the study end date (December 31, 2003) were censored at this time or at date of last follow-up (ie, last known contact). Seventy percent of patients had a follow-up date within 1 year of the study end date, whereas 80% had a follow-up within 2 years. EBV status did not differ for patients with and without follow-up information within 2 years.
Differences in OS and DSS Kaplan-Meier survival curves between EBV-positive and EBV-negative patients were compared statistically using the log-rank test. Kaplan-Meier curves were stratified by 15-year age groups (15 to 29, 30 to 44, 45 to 59, 60 to 74, and 75+ years), sex, race/ethnicity (non-Hispanic white v other), histologic subtype, presence of B symptoms, stage of disease, and initial treatment. Given our previously noted deficit in OS for EBV-positive HL only among women more than 45 years old1 and other survival differences by age,1 we ultimately conducted survival analyses separately for children (0 to 14 years old), younger adults (15 to 44 years old), and older adults (45 years old and older). Because of small numbers, lymphocyte-rich and lymphocyte-depleted histologic subtypes were combined for multivariate analyses.
To evaluate the impact of EBV status on survival controlling for other prognostic factors, we used Cox proportional hazards regression to determine hazard ratios (HR), which estimate the instantaneous relative risk of death averaged over the entire time period, and associated 95% CIs. Multivariate analyses included variables significant at P < .15 in univariate models or with a priori hypotheses for inclusion (eg, age, race/ethnicity, and sex). In all models, the proportional hazards assumption was assessed by visual inspection of the survival curves (log [–log] of the survival distribution function by log [months]). Effect modification between covariates was examined by entering interaction terms into the multivariate model. All HRs presented were adjusted for age and race/ethnicity unless otherwise noted. All analyses were conducted using SAS version 9 software (SAS Institute Inc, Cary, NC).
RESULTS
Cohort Characteristics
Of the 922 patients, 67% were between the ages of 15 and 44 years at diagnosis, and 73% were of non-Hispanic white race/ethnicity. The group included nearly equal proportions of males and females. Seventy-four percent of the patients had the NS histologic subtype, and 61% had Ann Arbor stage I or II disease at diagnosis. As of the study end date, 74% of the cohort was alive.
Table 1 demonstrates that the proportion of all patients with EBV-positive tumors was 27%, but this rate varied by age (40% in children and older adults and 19% in young adults). Overall, EBV-positive patients were more likely than EBV-negative patients to be nonwhite and male, live in lower SES neighborhoods, and have the MC histologic subtype. In children, EBV-positive patients were less likely than EBV-negative patients to have stage IV disease or B symptoms. In younger adults, EBV-positive patients were somewhat more likely to have a later stage of disease (stage III/IV) at diagnosis. In older adults, EBV-positive patients were more likely to have B symptoms and somewhat more likely to have had a prior cancer than EBV-negative patients. The type of initial treatment did not differ significantly between EBV-positive and EBV-negative patients within groups defined by age (age 15 to 44 years or 45+ years) or stage at diagnosis (stage I/II or III/IV; data not shown). Vital status also differed by age and EBV status in children (deaths in 6% of EBV-positive and 30% of EBV-negative patients) and older adults (deaths in 74% of EBV-positive and 53% of EBV-negative patients) but not in young adults (deaths in 12% of both EBV-positive and EBV-negative patients).
Survival Characteristics
The relationship of EBV to both OS and DSS differed markedly across age group. For children, EBV-positive HL seemed to have better OS and DSS (all deaths were a result of HL) than EBV-negative HL, as shown in Figure 1 (log-rank test, P = .07) and by Cox regression analysis (HR = 0.18; 95% CI, 0.02 to 1.7). However, these findings were not statistically significant, and the small numbers of deaths in this age group precluded further multivariate analyses.
In young adults, EBV status was not significantly associated with OS (Fig 2A; log-rank test, P = .76) or DSS (Fig 2B; log-rank test, P = .40). Cox models confirmed that EBV-positive tumor status did not impact HL outcome for OS overall or in data stratified by the NS or MC histologic subtypes (Table 2) or for DSS (HR = 0.59; 95% CI, 0.25 to 1.4). EBV presence also did not significantly influence survival in patient subgroups defined by race/ethnicity, sex, stage at diagnosis, presence of B symptoms, initial treatment, or age (for 346 patients 15 to 29 years old: HR for OS = 1.3; 95% CI, 0.62 to 2.9; HR for DSS = 0.73; 95% CI, 0.21 to 2.5; for 271 patients 30 to 44 years old: HR for OS = 0.37; 95% CI, 0.13 to 1.1; HR for DSS = 0.40; 95% CI, 0.12 to 1.4), although data were suggestive of better OS for 30 to 44 year olds.
In older adults, OS was substantially poorer for patients with EBV-positive than EBV-negative tumors (Fig 3A; log-rank test, P < .01). This deficit persisted irrespective of sex (men: HR = 1.4; 95% CI, 0.93 to 2.3; women: HR = 1.6; 95% CI, 1.0 to 2.7), stage at diagnosis, initial treatment, sex, race/ethnicity, and presence of B symptoms (data not shown). It also persisted across smaller age groups, albeit less so for the oldest age group (for 111 patients 45 to 59 years old: HR = 1.9; 95% CI, 1.0 to 3.7; for 105 patients 60 to 74 years old: HR = 1.6; 95% CI, 0.97 to 2.5; for 53 patients 75 to 96 years old: HR = 1.3; 95% CI, 0.70 to 2.4). Results were similar for DSS (Fig 3B) because most of the survival differences by EBV status were a result of HL-related deaths. In data stratified by histologic subtype, the deficit in OS was apparent for patients with NS (Fig 4A; log-rank test, P < .01) but not for patients with MC (Fig 4B; log-rank test, P = .80). Survival was better for patients with EBV-negative NS (76% at 50 months) than for patients with EBV-positive NS, EBV-positive MC, or EBV-negative MC, in whom the survival rate was lower and similar (approximately 50% at 50 months). Cox models (Table 3) confirmed that the hazard of death from any cause was elevated in EBV-positive compared with EBV-negative older adults, reflecting a 2.5-fold increase in NS patients. Similar results occurred for DSS, with HRs (adjusted for age, race/ethnicity, sex, histologic subtype, B symptoms, and disease stage) of 1.8 (95% CI, 1.1 to 2.9) overall, 2.4 (95% CI, 1.2 to 5.1) for NS, and 1.5 (95% CI, 0.6 to 3.7) for MC.
Among patients with nodular LP HL, 66% were male, 60% were white, and 10% were EBV positive. Over the follow-up period, 22% died; OS was suggestively poorer for EBV-positive than EBV-negative patients (log-rank test, P = .09; HR = 2.0; 95% CI, 0.49 to 8.5).
DISCUSSION
This population-based study has confirmed that EBV presence in HL tumor cells is a meaningful molecular marker of outcome after HL in both males and females, independent of other factors, but that its influence varies by patient age at diagnosis and HL histologic subtype. In children, EBV-positive status was suggestively associated with more favorable survival than EBV-negative status. In young adults, EBV did not significantly affect HL outcome, although the data suggested a possible protective effect in patients aged 30 to 44 years at diagnosis. In contrast, in older adults, EBV presence nearly doubled the risk of overall and HL-specific mortality, even after adjustment for age, stage at diagnosis, and other factors, but this effect was limited to patients with NS histologic subtype. Overall, our findings suggest that tumor cell EBV plays a complex role in the progression of HL but might serve as a useful indicator of prognosis in some groups, including older adults with NS and, possibly, children.
The relationship of EBV presence in HL tumor cells to survival after HL has been evaluated in 19 published studies, with varied results (Table 4). 11-30 The range of findings is likely attributable in part to differences in patient populations across studies, infrequent consideration of age, limited statistical power as a result of small sample sizes, and differing outcome measures (OS, DSS, or disease/failure-free survival). However, our current findings are consistent with those of studies using unselected patient groups and acknowledging age effects. The two prior studies looking at children reported a favorable effect of EBV-positive HL on survival,16,26 as we have noted here. Prior studies composed primarily of young adults found no EBV-related difference in survival,12,14,21,41 whereas three studies reported favorable OS among patients with EBV-positive tumors.23,24,26 Flavell et al18 found a trend towards longer disease-free survival in patients 15 to 34 years of age, and Glavina-Durdov et al19 observed a longer disease-free survival in EBV-positive patients under 35 years of age with an opposite finding in older adults. These findings are generally consistent with our null associations in young adults and suggestive observations for persons aged 30 to 44 years.18,19 Two prior population-based studies of adults greater than age 60 years found poorer OS15,27 and DSS15 in EBV-positive compared with EBV-negative patients, which is similar to our finding in adults older than 45 years.
However, few studies have considered the modifying effect of histologic subtype for EBV presence on HL outcome. Five studies reported no differences across subtypes but did not present their findings in detail,11,12,17,29,30 and one study combined NS and MC subtypes.19 No prior study presented data stratified jointly by age and histologic subtype. We have shown a significant interaction of age, EBV status, and histologic subtype with survival after HL, with poorer outcome for adults diagnosed at greater than age 44 years with EBV-positive NS, which is a group that comprises one third of all older HL patients.42 The presence of EBV in HL has been one of the criteria used for therapies such as EBV-specific cell infusion.43 Our findings suggest that detection of EBV in tumor cells might be relevant to guiding additional therapies, perhaps including high-dose therapy, in the older age group in whom the presence of EBV confers a poor prognosis. However, comorbidities and poor tolerance to cytotoxic therapies may limit the applicability of cytotoxic intensification in this age group.
EBV could promote tumor maintenance through a variety of mechanisms, including the same pathways it uses to persist in healthy hosts, namely increasing host cell proliferation, decreasing apoptosis, and eluding the immune system in part by downregulating HLA-mediated presentation of viral peptides. The variation in impact of tumor cell EBV on HL outcomes in patient groups defined by age and histology could indicate that these groups represent biologically distinct disease entities, as suggested by ample epidemiologic evidence6 and differences in EBV-specific cellular immunity in young and older adults.44 The poorer general immunocompetence of older patients may facilitate faster growth of an EBV-positive tumor than what might occur in children or young adults; in younger patients, EBV antigens might stimulate more robust anti-EBV or antitumor immune responses.45,46 Viral gene expression might be expected to be immunogenic at all ages because EBV-specific cytotoxic T cells can be isolated from most HL patients.47 However, EBV-positive HL tumor cells may escape immunosurveillance.48 LMP1 expression may alter immune response insofar as the viral protein induces expression of interleukin (IL) -10 or IL-6, which are cytokines that perturb the cellular immune response.49,50 Thus, high concentrations of IL-10 within the tumor mass may suppress EBV-specific cytotoxicity sufficiently to allow malignant cells to thrive, particularly given age-related declines in immunocompetence. Alternatively, cytokine imbalances could have other detrimental influences on HL survival.51 These mechanisms could be explored with more basic research addressing EBV-specific immune responses and the ways they are modified by shifting T-helper cytokine balance or other aspects of immunocompetence that might differ between young and older adult HL patients.
This analysis had the advantages of using a large group of HL patients of all ages identified from a population-based cancer registry with uniformly collected demographic and clinical data, expert histopathologic rereview, and relatively long follow-up. However, because we could not obtain tumor blocks for all eligible patients, patients in our analysis differed from those who were excluded. Those patients who were included were more likely to be female because the earlier study of women52 occurred when specimen retrieval was easier. Among children, included patients, compared with children who were not included, were more likely to have HL not otherwise specified (3% v 14%, respectively) and higher SES (38% v 26% in the highest quintile, respectively). Among older adults, included versus nonincluded patients were more like to be non-Hispanic white (75% v 65%, respectively) and to have characteristics associated with better survival (NS histologic subtype, earlier stage disease, and no B symptoms), in part, because we lacked patient consent for block release from deceased patients. These differences are likely to have affected our statistical power by reducing our sample size and may have biased our findings toward the null, especially in older adults in whom differences were most marked. Our study lacked information on disease recurrence, detailed treatment, and comorbid conditions, which are not collected by the cancer registry; however, EBV tumor cell status has been shown to have similar associations with disease-free and failure-free survival and with OS and/or DSS in most14,15,23,25,26 but not all18,19,21 studies with more detailed clinical information. Moreover, we found no association between first course of treatment and EBV presence when considering age at diagnosis and stage of disease, which is similar to a Swedish study of uniformly treated patients.15
Our findings that EBV in HL tumors decreased OS and DSS in older NS patients while conferring a survival advantage in children and possibly young adults indicate that EBV histochemical stains performed in these patient groups may assist in prognosis and clinical management decisions. The findings also support further investigation of EBV as a prognostic marker for HL. Understanding the mechanisms behind these associations will not only improve our understanding of the complexities of HL pathogenesis but may also be helpful in developing new therapeutic strategies targeting EBV directly or in designing better conventional regimens. However, future studies should include larger numbers of children and young adults to study the joint effects of age and EBV status with other prognostic factors and should include detailed clinical information to identify subsets of patients who respond differentially to therapy and, thus, to enhance the therapeutic applicability of these observations.
Authors' Disclosures of Potential Conflicts of Interest
Margaret L. GulleyUS Public Health Funded Project (NIH) (B)Richard F. AmbinderNCI (C)
NOTES
Supported by National Cancer Institute (NCI) Grants No. NCI R29 CA-50381 (S.L.G.), NCI RO3 CA-63245 (S.L.G., R.F.A.), and N01-CN-65107 (S.L.G.). The collection of Cancer incidence data used in this study was supported by the California Department of Health Services as part of the statewide cancer reporting program mandated by California Health and Safety Code Section 103885; the National Cancer Institute's Surveillance, Epidemiology and End Results Program under contract N01-PC-35136 awarded to the Northern California Cancer Center, contract N01-PC-35139 awarded to the University of Southern California, and contract N02-PC-15105 awarded to the Public Health Institute; and the Centers for Disease Control and Prevention's Nationals Program of Cancer Registries, under agreement #U55/CCR921930-02 awarded to the Public Health Institute. The ideas and opinions expressed herein are those of the authors, and endorsement by the state of California, Department of Health Services, the National Cancer Institute, and the Centers for Disease Control and Prevention or their contractors and subcontractors is not intended, nor should it be inferred.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Clarke CA, Glaser SL, Prehn AW: Age-specific survival after Hodgkin's disease in a population-based cohort (United States). Cancer Causes Control 12:803-812, 2001
Kennedy BJ, Loeb V Jr, Peterson V, et al: Survival in Hodgkin's disease by stage and age. Med Pediatr Oncol 20:100-104, 1992
Kennedy BJ, Fremgen AM, Menck HR: Hodgkin's disease survival by stage and age. J Am Geriatr Soc 48:315-317, 2000
Walker A, Schoenfeld ER, Lowman JT, et al: Survival of the older patient compared with the younger patient with Hodgkin's disease: Influence of histologic type, staging, and treatment. Cancer 65:1635-1640, 1990
Ries LAG, Elsner MP, Kosary CL, et al: (eds): SEER Cancer Statistics Review, 1973-1999. Bethesda, MD, National Cancer Institute, 2002
Mueller N: Hodgkin's disease, in Schottenfeld D, Fraumeni JF Jr (eds): Cancer Epidemiology and Prevention (ed 2). New York, NY, Oxford University Press, 1996, pp 893-919
Mueller N, Grufferman S: The epidemiology of Hodgkin's disease, in Mauch PM, Armitage JO, Diehl V, et al (eds): Hodgkin's Disease. Philadelphia, Lippincott Williams & Wilkins, 1999, pp 61-77
Flavell KJ, Biddulph JP, Constandinou CM, et al: Variation in the frequency of Epstein-Barr virus-associated Hodgkin's disease with age. Leukemia 14:748-753, 2000
Glaser SL, Jarrett RF: The epidemiology of Hodgkin's disease. Baillieres Clin Haematol 9:401-416, 1996
Jarrett RF: Risk factors for Hodgkin's lymphoma by EBV status and significance of detection of EBV genomes in serum of patients with EBV-associated Hodgkin's lymphoma. Leuk Lymphoma 44:S27-S32, 2003 (suppl 3)
Armstrong AA, Lennard A, Alexander FE, et al: Prognostic significance of Epstein-Barr virus association in Hodgkin's disease. Eur J Cancer 30A:1045-1046, 1994
Axdorph U, Porwit-MacDonald A, Sjoberg J, et al: Epstein-Barr virus expression in Hodgkin's disease in relation to patient characteristics, serum factors and blood lymphocyte function. Br J Cancer 81:1182-1187, 1999
Clarke CA, Glaser SL, Dorfman RF, et al: Epstein-Barr virus and survival after Hodgkin's disease in a population-based series of women. Cancer 91:1579-1587, 2001
Enblad G, Sandvej K, Lennette E, et al: Lack of correlation between EBV serology and presence of EBV in the Hodgkin and Reed-Sternberg cells of patients with Hodgkin's disease. Int J Cancer 72:394-397, 1997
Enblad G, Sandvej K, Sundstrom C, et al: Epstein-Barr virus distribution in Hodgkin's disease in an unselected Swedish population. Acta Oncol 38:425-429, 1999
Engel M, Essop MF, Close P, et al: Improved prognosis of Epstein-Barr virus associated childhood Hodgkin's lymphoma: Study of 47 South African cases. J Clin Pathol 53:182-186, 2000
Fellbaum C, Hansmann ML, Niedermeyer H, et al: Influence of Epstein-Barr virus genomes on patient survival in Hodgkin's disease. Am J Clin Pathol 98:319-323, 1992
Flavell KJ, Billingham LJ, Biddulph JP, et al: The effect of Epstein-Barr virus status on outcome in age- and sex-defined subgroups of patients with advanced Hodgkin's disease. Ann Oncol 14:282-290, 2003
Glavina-Durdov M, Jakic-Razumovic J, Capkun V, et al: Assessment of the prognostic impact of the Epstein-Barr virus-encoded latent membrane protein-1 expression in Hodgkin's disease. Br J Cancer 84:1227-1234, 2001
Herling M, Rassidakis GZ, Medeiros LJ, et al: Expression of Epstein-Barr virus latent membrane protein-1 in Hodgkin and Reed-Sternberg cells of classical Hodgkin's lymphoma: Associations with presenting features, serum interleukin 10 levels, and clinical outcome. Clin Cancer Res 9:2114-2120, 2003
Krugmann J, Tzankov A, Gschwendtner A, et al: Longer failure-free survival interval of Epstein-Barr virus-associated classical Hodgkin's lymphoma: A single-institution study. Mod Pathol 16:566-573, 2003
Levy A, Diomin V, Gopas J, et al: Hodgkin's lymphoma in the Bedouin of southern Israel: Epidemiological and clinical features. Isr Med Assoc J 2:501-503, 2000
Montalban C, Abraira V, Morente M, et al: Epstein-Barr virus-latent membrane protein 1 expression has a favorable influence in the outcome of patients with Hodgkin's disease treated with chemotherapy. Leuk Lymphoma 39:563-572, 2000
Morente MM, Piris MA, Abraira V, et al: Adverse clinical outcome in Hodgkin's disease is associated with loss of retinoblastoma protein expression, high Ki67 proliferation index, and absence of Epstein-Barr virus-latent membrane protein 1 expression. Blood 90:2429-2936, 1997
Murray PG, Billingham LJ, Hassan HT, et al: Effect of Epstein-Barr virus infection on response to chemotherapy and survival in Hodgkin's disease. Blood 94:442-447, 1999
Naresh KN, Johnson J, Srinivas V, et al: Epstein-Barr virus association in classical Hodgkin's disease provides survival advantage to patients and correlates with higher expression of proliferation markers in Reed-Sternberg cells. Ann Oncol 11:91-96, 2000
Stark GL, Wood KM, Jack F, et al: Hodgkin's disease in the elderly: A population-based study. Br J Haematol 119:432-440, 2002
Vassallo J, Metze K, Traina F, et al: Expression of Epstein-Barr virus in classical Hodgkin's lymphomas in Brazilian adult patients. Haematologica 86:1227-1228, 2001
Vassallo J, Metze K, Traina F, et al: The prognostic relevance of apoptosis-related proteins in classical Hodgkin's lymphomas. Leuk Lymphoma 44:483-488, 2003
Vestlev PM, Pallesen G, Sandvej K, et al: Prognosis of Hodgkin's disease is not influenced by Epstein-Barr virus latent membrane protein. Int J Cancer 50:670-671, 1992
Clarke CA, Glaser SL, Keegan TH, et al: Neighborhood socioeconomic status and Hodgkin's lymphoma incidence in California. Cancer Epidemiol Biomarkers Prev 14:1441-1447, 2005
Yost K, Perkins C, Cohen R, et al: Socioeconomic status and breast cancer incidence in California for different race/ethnic groups. Cancer Causes Control 12:703-711, 2001
Glaser SL, Dorfman RF, Clarke CA: Expert review of the diagnosis and histologic classification of Hodgkin's disease in a population-based cancer registry: Interobserver reliability and impact on incidence and survival rates. Cancer 92:218-224, 2001
Diehl V, Sextro M, Franklin J, et al: Clinical presentation, course, and prognostic factors in lymphocyte-predominant Hodgkin's disease and lymphocyte-rich classical Hodgkin's disease: Report from the European Task Force on Lymphoma Project on Lymphocyte-Predominant Hodgkin's Disease. J Clin Oncol 17:776-783, 1999
Pileri SA, Ascani S, Leoncini L, et al: Hodgkin's lymphoma: The pathologist's viewpoint. J Clin Pathol 55:162-176, 2002
Gulley ML, Glaser SL, Craig FE, et al: Guidelines for interpreting EBER in situ hybridization and LMP1 immunohistochemical tests for detecting Epstein-Barr virus in Hodgkin's disease. Am J Clin Pathol 117:259-267, 2002
Murray PG, Constandinou CM, Crocker J, et al: Analysis of major histocompatibility complex class I, TAP expression, and LMP2 epitope sequence in Epstein-Barr virus-positive Hodgkin's disease. Blood 92:2477-2483, 1998
Glaser SL, Keegan THM, Clarke CA, et al: Exposure to childhood infections and risk of EBV-defined Hodgkin's lymphoma in women. Int J Cancer 115:599-605, 2005
Clarke C, Glaser SL: Population-based surveillance of HIV-associated cancers: Utility of cancer registry data. J Acquir Immune Defic Syndr 36:1083-1091, 2004
Glaser SL, Clarke CA, Gulley ML, et al: Population-based patterns of human immunodeficiency virus-related Hodgkin lymphoma in the Greater San Francisco Bay Area, 1988-1998. Cancer 98:300-309, 2003
Herrmann K, Niedobitek G: Lack of evidence for an association of Epstein-Barr virus infection with breast carcinoma. Breast Cancer Res 5:R13-R17, 2003
Surveillance Epidemiology and End Results (SEER) Program: SEER*Stat Database: Incidence—SEER 9 Regs Public-Use, Nov 2003 Sub (1973-2001). Bethesda, MD, National Cancer Institute, Division of Cancer Control and Population Sciences, Surveillance Research Program, Cancer Statistics Branch, November, 2003
Rooney CM, Bollard C, Huls MH, et al: Immunotherapy for Hodgkin's disease. Ann Hematol 81:S39-S42, 2002 (suppl 2)
Gandhi MK, Tellam JT, Khanna R: Epstein-Barr virus-associated Hodgkin's lymphoma. Br J Haematol 125:267-281, 2004
Armstrong AA, Alexander FE, Cartwright R, et al: Epstein-Barr virus and Hodgkin's disease: Further evidence for the three disease hypothesis. Leukemia 12:1272-1276, 1998
Frisan T, Sjoberg J, Dolcetti R, et al: Local suppression of Epstein-Barr virus (EBV)-specific cytotoxicity in biopsies of EBV-positive Hodgkin's disease. Blood 86:1493-1501, 1995
Roskrow MA, Suzuki N, Gan Y, et al: Epstein-Barr virus (EBV)-specific cytotoxic T lymphocytes for the treatment of patients with EBV-positive relapsed Hodgkin's disease. Blood 91:2925-2934, 1998
Rickinson AB, Kieff E: Epstein-Barr virus, in Fields BN, Knipe DM, Howley PM, et al (eds): Field's Virology (ed 3). Philadelphia, PA, Lippincott-Raven Publishers, 1996, pp 2397-2446
Herbst H, Foss HD, Samol J, et al: Frequent expression of interleukin-10 by Epstein-Barr virus-harboring tumor cells of Hodgkin's disease. Blood 87:2918-2929, 1996
Herbst H, Samol J, Foss HD, et al: Modulation of interleukin-6 expression in Hodgkin and Reed-Sternberg cells by Epstein-Barr virus. J Pathol 182:299-306, 1997
Seymour JF, Talpaz M, Hagemeister FB, et al: Clinical correlates of elevated serum levels of interleukin 6 in patients with untreated Hodgkin's disease. Am J Med 102:21-28, 1997
Glaser SL, Clarke CA, Stearns CB, et al: Age variation in Hodgkin's disease risk factors in older women: Evidence from a population-based case-control study. Leuk Lymphoma 42:997-1004, 2001(Theresa H.M. Keegan, Sall)