Abstract
We evaluated associations between Epstein–Barr virus (EBV) antibody levels and precancerous gastric lesions (chronic atrophic gastritis, intestinal metaplasia, and dysplasia) in 183 subjects from Linqu, China. Immunoglobulin G antibody titers to EBV nuclear antigen (EBNA) and viral capsid antigen (VCA) were determined by two‐fold serial dilution using immunofluorescence assays. Histological progression and regression were assessed by gastroscopic examination at the time of phlebotomy and at follow up 2 years later. Antibody titers did not differ significantly among histological diagnoses determined at the time of phlebotomy. However, subjects with dysplasia at follow up had significantly higher geometric mean antibody titers for both anti‐VCA and anti‐EBNA. Subjects with greater than median antibody levels were more likely to progress between examinations, especially in the subgroup with intestinal metaplasia at the time of phlebotomy (odds ratios [95% confidence intervals]: 5.7 [1.6–20] for anti‐EBNA ≥1:320; 3.8 [1.0–15] for anti‐VCA ≥1:640). Our findings suggest a possible role for EBV reactivation at an early phase of gastric carcinogenesis. (Cancer Sci 2008; 99: 350–354)
With an estimated 934 000 incident cases and 700 000 deaths worldwide, gastric carcinoma (GC) is the fourth most common cancer, second only to lung cancer in cancer‐related mortality.( 1 ) GC rates vary widely throughout the world with the highest rates occurring in eastern Asia. Differences in these rates have been attributed to differences in diet and environmental exposures.( 2 )
The pathogenesis of gastric carcinoma is a multistage process affecting the gastric mucosa.( 3 ) For at least intestinal‐type GC, there is thought to be a histological progression of precancerous lesions leading up to GC, beginning with superficial gastritis (SG) and advancing through chronic atrophic gastritis (CAG), intestinal metaplasia (IM), dysplasia, and finally to GC. More abnormal lesions are associated with an increased risk of GC. In Linqu, China, individuals diagnosed with dysplasia are more likely than those with SG or CAG to develop GC after 5 years.( 4 )
Infection and colonization of the stomach by the bacterium Helicobacter pylori is an important risk factor in the development of most GC, especially non‐cardia GC.( 5 , 6 , 7 ) H. pylori infection is thought to cause the development of CAG and IM,( 8 ) and the eradication of H. pylori results in a reduction in the risk of precancerous gastric lesions.( 9 ) A history of H. pylori infection has been found in most GC cases.( 7 , 10 )
Other infectious agents may also have a role in GC. In particular, the gammaherpesvirus Epstein–Barr virus (EBV) has been implicated by both molecular and serological evidence. EBV nucleic acids are detectable by polymerase chain reaction or in situ hybridization in a significant percentage of GC tumors from many regions of the world, including 16% of GC tumors in North America,( 11 ) 7% in Japan,( 12 , 13 ) 9% in Russia,( 14 ) and 10% in Kazakhstan.( 15 ) In a prospective study of Japanese Americans, subjects who developed EBV‐positive GC had significantly increased antibody titers to EBV viral capsid antigen (VCA) prior to diagnosis.( 16 ) Moreover, in a case‐control study from Japan, patients with EBV‐positive gastric cancer had higher levels of anti‐VCA and anti‐early antigen (EA) antibody than those with EBV‐negative tumors, but both cancer groups had higher antibody levels than healthy controls.( 17 )
Previous studies have investigated the association of EBV infections and precancerous lesions of the stomach with mixed results. In some studies, EBV has been detected in CAG and IM lesions.( 18 , 19 ) Other studies detected EBV uniformly in all cells of EBV‐positive GC tumors and not in the surrounding normal epithelium( 20 ) or precancerous lesions.( 21 ) In one case report of a stem‐cell transplant patient, EBV reactivation in the setting of graft‐versus‐host disease with gastric mucosal damage led to EBV‐positive gastric dysplasia and rapid initiation and progression of GC.( 22 )
The present study examines the associations between EBV antibody titers and the progression of precancerous lesions of the stomach in a population at high risk of GC. We investigated whether elevated levels of antibodies to either the EBV VCA or EBV nuclear antigen (EBNA) were associated with more severe abnormalities of gastric histology.
Materials and Methods
Study design. The selection and histological screening for the Linqu observation study is described in detail elsewhere.( 4 , 23 ) Briefly, the full study population consists of 3399 individuals, aged 35–65 years, enrolled in 1989 from 14 villages of Linqu County, representing 83% of the eligible population. The population comes from a poor, rural region of China and subjects share low social economic status with similar lifestyles. Individuals underwent gastroscopic examinations during which mucosal biopsies were taken from seven standard sites as well as any other suspicious areas. These biopsies were analyzed at the Beijing Institute for Cancer Research by three pathologists. Nearly all individuals had some level of histological abnormality. Subjects were classified based on the most abnormal histological sample analyzed for that individual. Abnormalities ranging from least to most abnormal were SG (2% of individuals), CAG (45%), IM (33%), and dysplasia (20%) using pathological criteria described previously.( 4 , 23 ) For the current analysis, SG and CAG have been grouped together (SG/CAG).
The present study examines the associations of EBV antibody titers with gastric histology in two subgroups of the original 3399 study participants. Seventy‐one individuals originally diagnosed with SG/CAG from a single village (cohort 1) and 112 individuals originally diagnosed with gastric dysplasia from all 14 villages (cohort 2) were selected based on specimen availability. Three years after cohort enrollment, serum was obtained from each individual in 1992, representing the baseline year for the current study.
All but 2 of these 183 individuals underwent a gastroscopic examination at baseline in 1992 and 161 subjects received a 2‐year follow‐up examination in 1994. On each examination, individuals were classified according to the most abnormal biopsy among seven standard and other suspicious areas, as described above. At the time of phlebotomy, 19 (28%) of 69 subjects diagnosed as CAG at enrollment had progressed to IM or dysplasia, whereas 64 (57%) of 112 dysplasia subjects had regressed to either CAG or IM. On the follow‐up examination, 37 subjects had more abnormal histology compared to baseline, whereas 34 individuals had an improved histology.
Epstein–Barr virus antibody levels were determined by immunofluorescence assays on serial, two‐fold dilutions of serum drawn at baseline. Immunoglobulin (Ig) G antibodies to EBNA, latent membrane protein (LMP)‐2, EA, and VCA and IgA antibodies to EA and VCA were detected using protocols described previously.( 16 , 24 )
Statistical methods. Variations of geometric mean antibody titers by age, sex, and cohort membership were examined by trend tests and t‐tests as appropriate. Geometric mean antibody titers by histological diagnoses were adjusted for age, sex, and cohort. Trend tests for variation with histological diagnosis assumed equal‐spaced categories and were adjusted for age, sex, and cohort membership. Multivariable logistic regression was used to investigate associations of elevated anti‐VCA and anti‐EBNA antibody titers with histological progression (i.e. to more abnormal) or regression (i.e. to less abnormal) between the two examinations. Anti‐VCA and anti‐EBNA IgG levels were classified as elevated if titers were higher than the median values (≥1:640 for anti‐VCA, ≥1:320 for anti‐EBNA). Fisher's combined P method was used to determine overall significance of individual trends of cohort 1 and cohort 2 geometric mean antibody titers with respect to follow‐up histology.( 25 ) All P‐values reported are two‐sided. Odds ratios (OR) were considered statistically significant if the 95% confidence interval (CI) excluded 1.0. Data analysis was carried out using Stata 9.2 (StataCorp LP).
Results
Table 1 summarizes age, sex, and follow‐up histology of the sampled population, by baseline histological diagnosis. The mean age at enrollment for the entire group was 48.8 years and 58% were men. Gastric histology was associated with both age and sex. Older individuals (P = 0.002, test for trend) and men (P = 0.03, test for trend) had more abnormal gastric lesions at baseline.
Table 1.
Selected characteristics by baseline diagnosis
| Characteristic | SG/CAG (n = 61) | IM (n = 67) | DYS (n = 53) |
|---|---|---|---|
| Mean age ± SD (years) | 45.7 ± 8.4 | 50.0 ± 9.1 | 50.8 ± 7.8 |
| Number of men (%) | 29 (47.5) | 40 (59.7) | 36 (67.9) |
| Histology worse at follow up | 18 of 55 | 19 of 61 | NA |
| Histology improved at follow up | NA | 12 of 61 | 22 of 43 |
DYS, dysplasia; IM, intestinal metaplasia; NA, not applicable; SG/CAG, superficial gastritis or chronic atrophic gastritis.
Several of the antibodies showed little variation in the tested population. Anti‐LMP2 antibodies were not detected in any individual. Only 22 individuals were positive for anti‐EA IgG antibodies, including five with SG/CAG, nine with IM, and eight with dysplasia at baseline (P = 0.49, χ2‐test for heterogeneity). Only 3 of 183 individuals were positive for anti‐VCA IgA antibodies, all diagnosed with IM at baseline, whereas no individual tested positive for anti‐EA IgA antibodies.
There was sufficient variability in the titers of the other two antibodies to be analyzed in further detail. The distribution of anti‐VCA IgG antibody titers was: 1:20 (n = 4), 1:40 (n = 1), 1:80 (n = 26), 1:160 (n = 25), 1:320 (n = 56), 1:640 (n = 32), 1:1280 (n = 30), 1:2560 (n = 5), and 1:5120 (n = 4). Anti‐ENBA antibody titers were: 1:5 (n = 1), 1:20 (n = 14), 1:40 (n = 8), 1:80 (n = 45), 1:160 (n = 41), 1:320 (n = 47), 1:640 (n = 24), and 1:1280 (n = 3).
Geometric mean anti‐VCA and anti‐EBNA antibody titers were examined by age, sex, cohort membership, baseline diagnosis, and follow‐up diagnosis (Table 2). Geometric mean antibody titers increased with age for anti‐VCA (P = 0.02, test for trend) but not for anti‐EBNA (P = 0.36, test for trend). Baseline antibody titers did not significantly differ by sex, cohort membership, or baseline histological diagnosis determined at the time of phlebotomy. However, subjects with dysplasia at follow up had higher baseline geometric mean antibody titers. Both anti‐VCA and anti‐EBNA were significantly higher with more advanced gastric lesions at follow up (P = 0.001 and P = 0.008, respectively, test for trend).
Table 2.
Geometric mean antibody titers and 95% confidence intervals (CI) by selected characteristics
| Characteristic | n | Anti‐VCA (95% CI) | Anti‐EBNA (95% CI) |
|---|---|---|---|
| Age (years) | |||
| 36–39 | 38 | 357 (257–496) | 189 (137–260) |
| 40–49 | 57 | 245 (183–327) | 156 (123–198) |
| 50–59 | 58 | 406 (304–544) | 140 (103–192) |
| 60–65 | 30 | 544 (367–808) | 159 (111–231) |
| Trend; P = 0.02 | Trend; P = 0.36 | ||
| Sex | |||
| Male | 107 | 320 (255–402) | 144 (118–176) |
| Female | 76 | 409 (329–509) | 178 (143–223) |
| T‐test; P = 0.14 | T‐test; P = 0.17 | ||
| Cohort | |||
| Cohort 1 † | 71 | 302 (242–377) | 157 (125–197) |
| Cohort 2 ‡ | 112 | 393 (314–490) | 158 (129–193) |
| T‐test; P = 0.12 | T‐test; P = 0.96 | ||
| Baseline diagnosis | |||
| SG/CAG | 61 | 370 (277–494) | 149 (113–197) |
| IM | 67 | 357 (304–419) | 159 (137–185) |
| DYS | 53 | 345 (253–470) | 170 (127–228) |
| Adjusted § trend; P = 0.79 | Adjusted § trend; P = 0.60 | ||
| Follow‐up diagnosis | |||
| SG/CAG | 53 | 224 (168–300) | 111 (83–148) |
| IM | 64 | 350 (299–410) | 157 (134–184) |
| DYS | 44 | 546 (398–748) | 223 (163–305) |
| Adjusted § trend; P = 0.001 | Adjusted § trend; P = 0.008 | ||
Cohort 1 is subjects with diagnosis of SG/CAG 3 years prior to baseline year of the current study.
Cohort 2 is subjects with diagnosis of DYS 3 years prior to baseline year of the current study.
Titers and trends have been adjusted for age, sex, and cohort membership.
DYS, dysplasia; EBNA, Epstein–Barr virus nuclear antigen; IM, intestinal metaplasia; SG/CAG, superficial gastritis or chronic atrophic gastritis; VCA, viral capsid antigen.
The variation in baseline antibody level between follow‐up histological diagnoses was further examined in each cohort separately (Table 3). Both anti‐VCA and anti‐EBNA increased across histological categories in each cohort, although the trends were significant only for cohort 2. Nevertheless, Fisher's combined tests for trends in the two cohorts considered together were significant for both anti‐VCA (P = 0.005) and anti‐EBNA (P = 0.04).
Table 3.
Geometric mean antibody titers and 95% confidence intervals (CI) by cohort and follow‐up diagnosis
| n | Anti‐VCA (95% CI) | Anti‐EBNA (95% CI) | |
|---|---|---|---|
| Cohort 1 † | |||
| SG/CAG | 45 | 250 (196–319) | 145 (111–190) |
| IM | 15 | 357 (262–485) | 179 (127–251) |
| DYS | 4 | 508 (274–944) | 219 (111–434) |
| Adjusted § trend; P = 0.05 | Adjusted § trend; P = 0.30 | ||
| Cohort 2 ‡ | |||
| SG/CAG | 8 | 207 (121–354) | 88 (52–147) |
| IM | 49 | 330 (256–426) | 133 (104–170) |
| DYS | 40 | 527 (378–736) | 201 (146–276) |
| Adjusted § trend; P = 0.01 | Adjusted § trend; P = 0.02 | ||
| Fisher's combined test; P = 0.005 | Fisher's combined test; P = 0.04 | ||
Cohort 1 is subjects with diagnosis of SG/CAG 3 years prior to baseline year of the current study.
Cohort 2 is subjects with diagnosis of DYS 3 years prior to baseline year of the current study.
Titers and trends have been adjusted for age and sex.
DYS, dysplasia; EBNA, Epstein–Barr virus nuclear antigen; IM, intestinal metaplasia; SG/CAG, superficial gastritis or chronic atrophic gastritis; VCA, viral capsid antigen.
Anti‐EBV antibody levels were associated with changes in gastric histology between the baseline and follow‐up examinations. Taking both cohorts together, subjects with elevated anti‐EBNA antibodies were significantly more likely to progress between baseline and follow‐up examinations (OR = 2.8, 95% CI = 1.1–6.8). The association was more pronounced among subjects with IM at baseline (OR = 5.7, 95% CI = 1.6–20) than among those with SG/CAG (OR = 1.2, 95% CI = 0.3–4.6). Elevated anti‐VCA antibodies were associated with progression in the subgroup with IM at baseline (OR = 3.8, 95% CI = 1.0–15), but not in all subjects taken together (OR = 1.5, 95% CI = 0.6–3.9). Subjects with elevated anti‐EBNA or anti‐VCA antibodies were non‐significantly less likely to regress between examinations, especially among those with IM at baseline (Table 4).
Table 4.
Adjusted † odds ratios (OR) for histological progression or regression with high Epstein–Barr virus (EBV) antibody titers, by baseline histological diagnosis
| Progression during follow up | |||
|---|---|---|---|
| SG/CAG or IM (n = 116) | SG/CAG only (n = 55) | IM only (n = 61) | |
| OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| EBNA ≥1:320 | 2.8 (1.1–6.8) | 1.2 (0.3–4.6) | 5.7 (1.6–20) |
| VCA ≥1:640 | 1.5 (0.6–3.9) | 0.5 (0.1–2.3) | 3.8 (1.0–15) |
| Regression during follow up | |||
|---|---|---|---|
| IM or DYS (n = 104) | IM only (n = 61) | DYS only (n = 43) | |
| OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| EBNA ≥1:320 | 0.7 (0.3–1.7) | 0.4 (0.1–2.1) | 0.8 (0.2–2.9) |
| VCA ≥1:640 | 0.7 (0.3–1.7) | 0.5 (0.1–2.5) | 0.9 (0.2–3.7) |
OR adjusted for age, sex and cohort membership.
DYS, dysplasia; EBNA, Epstein–Barr virus nuclear antigen; IM, intestinal metaplasia; SG/CAG, superficial gastritis or chronic atrophic gastritis; VCA, viral capsid antigen.
Discussion
Epstein–Barr virus exposure generally results in a lifelong infection of B lymphocytes. Primary EBV infections are usually asymptomatic if acquired in childhood but can lead to the presentation of infectious mononucleosis in a minority of adolescents.( 26 ) Infection rates of EBV are very high worldwide and reach almost universal levels.( 27 )
Epstein–Barr virus is associated with several types of cancer, including Burkitt's lymphoma, nasopharyngeal carcinoma, Hodgkin's lymphoma, and B‐cell lymphomas in immunosuppressed patients.( 27 ) EBV infection is capable of transforming normal B cells into immortalized cell lines in vitro.( 28 ) EBV infection can disrupt the proper function of cell‐cycle regulatory proteins in vitro, such as p21 or Cdk228. This alteration affects the cellular processes and checkpoints responsible for regulating cell division, which is thought to be a major component of the carcinogenic effects of EBV.
In our study, we observed trends between EBV antibody titers and the presence of precancerous gastric lesions at follow‐up examination, 2 years after blood draw. These findings are consistent with the hypothesis that EBV plays a role in the formation of these lesions.
Further support for this hypothesis comes from evidence that EBV stimulates replication of gastric epithelial cells. EBV infection of an EBV‐negative gastric cancer cell line has been shown to induce secretion of insulin‐like growth factor (IGF)‐1, which acts as an autocrine growth factor in vitro. ( 29 ) Furthermore, IGF‐1 mRNA levels are significantly higher both in gastric carcinoma tissue and in adjacent mucosa as compared with tumor‐free tissue.( 30 ) Thus, EBV‐induced IGF‐1 overexpression may be an early lesion in the progression of gastric carcinogenesis.
The stronger associations of progression and regression in individuals diagnosed with IM at baseline suggest that IM lesions may be more susceptible to EBV replication status. This finding agrees with the previous observation that damaged gastric mucosa may be more susceptible to EBV infection and consequent initiation and progression of gastric carcinogenesis.( 22 )
The associations we observed between EBV and precancerous gastric lesions require confirmation in independent investigations. Further studies should address whether EBV contributes to histological progression equally in different populations. Individuals with compromised gastric mucosa or a suppressed immune system may be more susceptible to the virus’ effects.
High anti‐EBV antibody titers at baseline were associated with more advanced gastric histology at follow up, but not at baseline. This temporal specificity implies a transient effect (i.e. EBV reactivation) rather than constitutively high antibody is associated with development of gastric lesions (otherwise, contemporaneous levels would also be associated with gastric histology). Nevertheless, we did not have longitudinal antibody measurements to confirm this conjecture.
Under the most accepted model, EBV should be present in precancerous lesions in association with high antibody titers in those cases where the virus is contributing to progression. Alternatively, EBV reactivation might increase the risk of formation and progression for all precancerous lesions, in line with the previous observation that even EBV‐negative GC is associated with elevated antibody titers to EBV antigens.( 17 ) However, tissue testing would not distinguish these possibilities definitively. Precancerous lesions adjacent to EBV‐positive tumors may be EBV negative,( 18 , 19 , 21 ) but as precancerous lesions are not clonal, their EBV status may not be representative of the subsequent monoclonal tumor. Additionally, increased EBV antibody levels are associated with both EBV‐positive and EBV‐negative gastric cancer,( 17 ) suggesting a possible indirect mechanism of action for EBV in gastric carcinogenesis. Thus, EBV in situ results will not be conclusive evidence for or against a role for EBV in precancerous gastric lesions. Nevertheless, further studies evaluating the prevalence of EBV in various precancerous lesions would be informative, particularly if such studies were linked with EBV serology.
In summary, we found significant associations between anti‐EBV antibodies and precancerous gastric lesions. Our findings are consistent with an etiological role for EBV in the early stages of gastric carcinogenesis.
Acknowledgments
This work was supported by the Intramural Research Program of the NIH, National Cancer Institute. AJS is funded by the Cancer Prevention Fellowship Program, Office of Preventive Oncology, National Cancer Institute, National Institute of Health.
References
- 1. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005; 55: 74–108. [DOI] [PubMed] [Google Scholar]
- 2. McMichael AJ, McCall MG, Hartshorne JM, Woodings TL. Patterns of gastro‐intestinal cancer in European migrants to Australia: the role of dietary change. Int J Cancer 1980; 25: 431–7. [DOI] [PubMed] [Google Scholar]
- 3. Correa P, Haenszel W, Cuello C, Tannenbaum S, Archer M. A model for gastric cancer epidemiology. Lancet 1975; 2: 58–60. [DOI] [PubMed] [Google Scholar]
- 4. You WC, Li JY, Blot WJ et al . Evolution of precancerous lesions in a rural Chinese population at high risk of gastric cancer. Int J Cancer 1999; 83: 615–19. [DOI] [PubMed] [Google Scholar]
- 5. Parsonnet J, Friedman GD, Vandersteen DP et al . Helicobacter pylori infection and the risk of gastric carcinoma. N Engl J Med 1991; 325: 1127–31. [DOI] [PubMed] [Google Scholar]
- 6. Nomura A, Stemmermann GN, Chyou PH, Kato I, Perez‐Perez GI, Blaser MJ. Helicobacter pylori infection and gastric carcinoma among Japanese Americans in Hawaii. N Engl J Med 1991; 325: 1132–6. [DOI] [PubMed] [Google Scholar]
- 7. Ekstrom AM, Held M, Hansson LE, Engstrand L, Nyren O. Helicobacter pylori in gastric cancer established by CagA immunoblot as a marker of past infection. Gastroenterology 2001; 121: 784–91. [DOI] [PubMed] [Google Scholar]
- 8. Asaka M, Kato M, Kudo M et al . Atrophic changes of gastric mucosa are caused by Helicobacter pylori infection rather than aging: studies in asymptomatic Japanese adults. Helicobacter 1996; 1: 52–6. [DOI] [PubMed] [Google Scholar]
- 9. You WC, Brown LM, Zhang L et al . Randomized double‐blind factorial trial of three treatments to reduce the prevalence of precancerous gastric lesions. J Natl Cancer Inst 2006; 98: 974–83. [DOI] [PubMed] [Google Scholar]
- 10. Uemura N, Okamoto S, Yamamoto S et al . Helicobacter pylori infection and the development of gastric cancer. N Engl J Med 2001; 345: 784–9. [DOI] [PubMed] [Google Scholar]
- 11. Shibata D, Weiss LM. Epstein–Barr virus‐associated gastric adenocarcinoma. Am J Pathol 1992; 140: 769–74. [PMC free article] [PubMed] [Google Scholar]
- 12. Tokunaga M, Land CE, Uemura Y, Tokudome T, Tanaka S, Sato E. Epstein–Barr virus in gastric carcinoma. Am J Pathol 1993; 143: 1250–4. [PMC free article] [PubMed] [Google Scholar]
- 13. Imai S, Koizumi S, Sugiura M et al . Gastric carcinoma: monoclonal epithelial malignant cells expressing Epstein–Barr virus latent infection protein. Proc Natl Acad Sci USA 1994; 91: 9131–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Galetsky SA, Tsvetnov VV, Land CE et al . Epstein–Barr‐virus‐associated gastric cancer in Russia. Int J Cancer 1997; 73: 786–9. [DOI] [PubMed] [Google Scholar]
- 15. Alipov G, Nakayama T, Nakashima M et al . Epstein–Barr virus‐associated gastric carcinoma in Kazakhstan. World J Gastroenterol 2005; 11: 27–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Levine PH, Stemmermann G, Lennette ET, Hildesheim A, Shibata D, Nomura A. Elevated antibody titers to Epstein–Barr virus prior to the diagnosis of Epstein–Barr‐virus‐associated gastric adenocarcinoma. Int J Cancer 1995; 60: 642–4. [DOI] [PubMed] [Google Scholar]
- 17. Shinkura R, Yamamoto N, Koriyama C, Shinmura Y, Eizuru Y, Tokunaga M. Epstein–Barr virus‐specific antibodies in Epstein–Barr virus‐positive and ‐negative gastric carcinoma cases in Japan. J Med Virol 2000; 60: 411–16. [DOI] [PubMed] [Google Scholar]
- 18. Yanai H, Takada K, Shimizu N, Mizugaki Y, Tada M, Okita K. Epstein–Barr virus infection in non‐carcinomatous gastric epithelium. J Pathol 1997; 183: 293–8. [DOI] [PubMed] [Google Scholar]
- 19. Hirano A, Yanai H, Shimizu N et al . Evaluation of Epstein–Barr virus DNA load in gastric mucosa with chronic atrophic gastritis using a real‐time quantitative PCR assay. Int J Gastrointest Cancer 2003; 34: 87–94. [DOI] [PubMed] [Google Scholar]
- 20. Shibata D, Tokunaga M, Uemura Y, Sato E, Tanaka S, Weiss LM. Association of Epstein–Barr virus with undifferentiated gastric carcinomas with intense lymphoid infiltration: Lymphoepithelioma‐like carcinoma. Am J Pathol 1991; 139: 469–74. [PMC free article] [PubMed] [Google Scholar]
- 21. Zur Hausen A, Van Rees BP, Van Beek J et al . Epstein–Barr virus in gastric carcinomas and gastric stump carcinomas: a late event in gastric carcinogenesis. J Clin Pathol 2004; 57: 487–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Au WY, Pang A, Chan EC, Chu KM, Shek TW, Kwong YL. Epstein–Barr virus‐related gastric adenocarcinoma: an early secondary cancer post hemopoietic stem cell transplantation. Gastroenterology 2005; 129: 2058–63. [DOI] [PubMed] [Google Scholar]
- 23. You WC, Blot WJ, Li JY et al . Precancerous gastric lesions in a population at high risk of stomach cancer. Cancer Res 1993; 53: 1317–21. [PubMed] [Google Scholar]
- 24. Lennette ET, Rymo L, Yadav M et al . Disease‐related differences in antibody patterns against EBV‐encoded nuclear antigens EBNA 1, EBNA 2 and EBNA 6. Eur J Cancer 1993; 29A: 1584–9. [DOI] [PubMed] [Google Scholar]
- 25. Fisher RA. Statistical Methods for Research Workers, 4th edn. Edinburgh: Oliver and Boyd, 1932. [Google Scholar]
- 26. Young LS, Rickinson AB. Epstein–Barr virus: 40 years on. Nat Rev Cancer 2004; 4: 757–68. [DOI] [PubMed] [Google Scholar]
- 27. Proceedings of the IARC Working Group on the Evaluation of Carcinogenic Risks to Humans . Epstein–Barr virus Kaposi's Sarcoma herpesvirus/human herpesvirus 8 Lyon, France, 17–24 June 1997. IARC Monogr Eval Carcinog Risks Hum 1997; 70. [Google Scholar]
- 28. O’Nions J, Allday MJ. Deregulation of the cell cycle by the Epstein–Barr virus. Adv Cancer Res 2004; 92: 119–86. [DOI] [PubMed] [Google Scholar]
- 29. Iwakiri D, Eizuru Y, Tokunaga M, Takada K. Autocrine growth of Epstein–Barr virus‐positive gastric carcinoma cells mediated by an Epstein–Barr virus‐encoded small RNA. Cancer Res 2003; 63: 7062–7. [PubMed] [Google Scholar]
- 30. Zhao MD, Hu XM, Sun DJ, Zhang Q, Zhang YH, Meng W. Expression of some tumor associated factors in human carcinogenesis and development of gastric carcinoma. World J Gastroenterol 2005; 11: 3217–21. [DOI] [PMC free article] [PubMed] [Google Scholar]