Markers of Past Infection with Simian Virus 40 (SV40) and Risk of Incident Non-Hodgkin Lymphoma in a Maryland Cohort

1448 Cancer Epidemiology, Biomarkers & Prevention Markers of Past Infection with Simian Virus 40 (SV40) and Risk of Incident Non-Hodgkin Lymphoma in a Maryland Cohort

Dana E. Rollison,1,2 Kathy J. Helzlsouer,2 Neal A. Halsey,3 Keerti V. Shah,4 and Raphael P. Viscidi5 1Division of Cancer Prevention and Control, H. Lee Moffitt Cancer Center and Research Institute, Tampa, Florida and Departments of 2Epidemiology, 3International Health, 4Molecular Microbiology and Immunology, and 5Pediatrics, Johns Hopkins School of Medicine, Baltimore, Maryland

Abstract

Simian virus 40 (SV40) genome sequences have been detected for cross-reactivity with JC virus (JCV) and BK virus (BKV) in human non-Hodgkin lymphoma (NHL) tissues, and past through competitive inhibition assays. Associations between infection with SV40 may be a risk factor for NHL. We SV40 antibody seropositivity and NHL were estimated using conducted a population-based nested case-control study to conditional logistic regression. Whereas SV40 antibodies investigate the association between serum antibodies to SV40 were detected by VLP ELISA in 15% of cases and 10% of and incident NHL. Two research serum banks were estab- controls [matched odds ratio (OR), 1.97; 95% confidence lished in Washington County, MD, with >45,000 volunteers interval (95% CI), 1.03-3.76], the SV40 reactivity of 85% of the contributing blood samples collected in 1974 and 1989. SV40 antibody-positive sera was decreased by adsorption Incident cases of NHL diagnosed through 2002 (n = 170) with BKV and/or JCV VLPs. Antibodies specific for SV40 were identified among participants by linkage to population- (not cross-reactive) were identified in only 1.8% of cases and based cancer registries. Two controls were matched to each 1.6% of controls (OR, 1.51; 95% CI, 0.41-5.52). Our findings case (n = 340) on age, sex, and date of blood draw. Circulating suggest that past infection with SV40 is not associated with immunoglobulin G antibodies to SV40 were measured using an increased risk of developing NHL. (Cancer Epidemiol virus-like particle (VLP) ELISA. Positive samples were tested Biomarkers Prev 2005;14(6):1448–52)

Introduction

The incidence of non-Hodgkin lymphoma (NHL) has nearly The evidence linking SV40 infection to NHL is inconsis- doubled in the United States since the early 1970s (1). Viral tent. Among nine case series which investigated SV40 in infections have been considered as potential risk factors for tumor tissues, six reported the presence of SV40 sequences NHL (2-4), including infection with simian virus 40 (SV40) (5). in 10% to 43% of cases (9-14), but three failed to detect SV40 SV40 is a natural polyomavirus infection of the rhesus sequences from NHL tissues (15-17). Findings from two macaque. Millions of Americans were potentially exposed to cohort studies were negative: no increases in NHL incidence SV40 through accidentally contaminated polio vaccines dis- rates were observed among HIV-positive individuals in the tributed in the United States between 1955 and 1963, after United States born in years during which the polio vaccine which time, polio vaccines were free of SV40 (6). Levels of was contaminated with SV40 compared with individuals SV40 contamination varied across polio vaccine lots, and it is born later (18) or in a study of national incidence rates of unclear how many individuals were actually exposed and NHL in Denmark (19). A population-based case-control infected with SV40. SV40 has been shown to be oncogenic in study of NHL in San Francisco found no associations with several laboratory models (7) and can produce B-cell lympho- self-reported history of polio vaccination, regardless of the mas in Syrian golden hamsters (8). The T antigen, a protein year of vaccination or age at vaccination (20). A hospital- coded by the virus, complexes with and subsequently based case-control study conducted among U.S. Army inactivates tumor suppressor proteins p53 and pRb. Veterans observed no associations between past exposure to SV40-contaminated adenovirus vaccine and NHL, brain tumors, or mesothelioma (21). None of these four studies

Received 9/13/04; revised 12/2/04; accepted 1/20/05. incorporated biomarkers of SV40 exposure. A hospital-based Grant support: National Cancer Institute grants 5 UO1 CA086308 (Early Detection Research case-control study of NHL in Spain observed an inverse Network), AG18033 (Odyssey Cohort), and 2 T32 CA093-14-19 (D.E. Rollison) and Centers for association between B-cell lymphomas and the presence of Disease Control and Prevention contract 200-2002-00732. serum antibodies to SV40 (5.9% positive in cases versus The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. 9.5% controls; ref. 22). The inactivated polio vaccine was Section 1734 solely to indicate this fact. not used universally in Spain (22); thus, the magnitude of Note: These data were supplied in part by the Maryland State Cancer Registry, Department of opportunity for SV40 exposure in this study population is Health and Mental Hygiene, Baltimore, MD. unclear. A recent population-based case-control study in the Conflict of interest: Dr. K.V. Shah has consulted with pharmaceutical companies that manufactured polio vaccines. No funding was provided by industry or commercial United States observed no increased risk of NHL associated enterprises, and these parties had no role in the study design, data collection, analysis with serum antibodies to SV40 measured after NHL or interpretation, or writing the report. diagnosis (23). Disclaimer: The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial To investigate the association between past SV40 infection products, or organizations imply endorsement by the U.S. Government. and the subsequent risk of NHL, we conducted a population- Requests for reprints: Dana E. Rollison, H. Lee Moffitt Cancer Center and Research Institute, based case-control study of serum antibodies to SV40 and 12902 Magnolia Drive, Tampa, FL 33612. Phone: 813-745-6530; Fax: 813-632-1334. E-mail: [email protected] incident NHL, nested within two community-based cohort Copyright D 2005 American Association for Cancer Research. studies in Washington County, MD.

Cancer Epidemiol Biomarkers Prev 2005;14(6). June 2005 Downloaded from cebp.aacrjournals.org on September 30, 2021. © 2005 American Association for Cancer Research. Cancer Epidemiology, Biomarkers & Prevention 1449

Materials and Methods Previous studies have shown that there is serologic cross- reactivity between SV40 and the two related human poly- Study Population. A research serum bank was established in omaviruses, JC virus (JCV) and BK virus (BKV) (25, 26). Washington County, MD, with 23,951 specimens collected from Therefore, samples testing positive for SV40 by ELISA county residents in August through November, 1974. This underwent competitive inhibition (‘‘blocking’’) assays to assess campaign was entitled ‘‘CLUE’’ for the slogan, ‘‘Give us a clue serologic cross-reactivity with JCV and BKV. VLP proteins for to heart disease and cancer.’’ A second CLUE campaign was JCV and BKV were prepared as described above for SV40. conducted between May and November 1989 with the collection Study serum samples were diluted 1:400 in buffer containing 4 of 25,079 blood specimens from county residents. Participants Ag/mL of SV40, BKV, or JCV VLP protein or in buffer alone. In completed a brief baseline questionnaire at the time of blood preliminary experiments, this concentration of SV40 VLP donation. Serum (1974) and plasma (1989) from participants protein in the diluent was shown to result in near maximal were stored at 70jCto73jC. All participants consented to inhibition of SV40 antibody-positive macaque sera. After the use of their stored blood in future research studies. The incubation of serum samples for 1 hour at 37jC on SV40- protocol for the current study was approved by the Committee coated microtiter plates, the ELISA assay was completed as on Human Research, the institutional review board at the Johns described above. The percent inhibition of SV40 reactivity by Hopkins Bloomberg School of Public Health. each VLP was calculated as 1 Acompeting VLP / Abuffer control Cases of NHL occurring among CLUE cohort members 100. If z50% of the SV40 reactivity was blocked by JCV and/or through 2002 were identified by linkage to the Washington BKV, then the SV40 antibodies were considered cross-reactive. County Registry [International Classification of Diseases, Ninth ‘‘SV40-specific’’ reactivity was defined as SV40 reactivity that Revision (ICD-9) code 200 or 202], which has been maintained was competitively inhibited z50% by SV40 and <50% by both since 1958, linkage to the Maryland State Cancer Registry since JCV and BKV (23). 1992, and periodic review of death certificates. Cases were Quality Control. Each case and its two matched controls defined as participants of CLUE I, CLUE II, or both, who were were maintained in their matched set to ensure simultaneous county residents at the times of both blood donation and processing. Laboratory personnel were masked as to the case- subsequent diagnosis with NHL, where NHL was their first control status of each sample. Seven pooled sera samples and cancer diagnosis with the possible exceptions of nonmelanoma eight pooled plasma samples were masked and placed across skin cancer or cervical cancer in situ. NHL subtypes were matched sets to test interset reliability of the ELISA assay. classified using ICD-9/ICD-10 morphology codes: diffuse large Similarly, nine duplicate pairs of sera and eight duplicate pairs B-cell lymphomas (ICD 9680/3, 9682/3, 9684/3, and 9680), of plasma were placed within matched sets to assess intraset follicular B-cell lymphomas (ICD 9675/3, 9690/3, 9691/3, reliability, each pair drawn from a cohort member not 9695/3, 9696/3, 9698/3, and 9690), T-cell lymphomas (ICD included in the study. Across sets, seven of seven pooled sera 9700/3), others/not specified (ICD 9590/3, 9591/3, 9592/3, samples were consistently negative for SV40 antibodies based 9593/3, 9595/3, 9670/3, 9672/3, 9673/3, 9686/3, 9687/3, on the absorbance value cut point of 0.1, whereas eight of eight 9694/3, 9699/3, 9711/3, 9823/3, 9940/3, 9533/1, 9693/3, and pooled plasma samples were consistently positive. Among the 9685/3). Cases (n = 172) were identified. Two cases were 17 duplicate pairs of sera or plasma, all but two samples were excluded due to inadequate amounts of available serum. negative for SV40 antibodies, and the two samples that tested Two controls were matched to each case on sex, race, age positive were from the same duplicate pair. In addition, seven within 1 year, date of blood draw within 2 weeks, freeze/thaw aliquots of SV40 antibody-positive monkey sera served as status of the serum, and participation in CLUE I, CLUE II, or masked positive controls; all of them showed high absorbance both. Controls were residents of Washington County at the values of >1.0 (mean absorbance value = 1.72, SD = 0.33). time of blood donation and not known to have died or developed cancer (except for possibly nonmelanoma skin Statistical Methods. Baseline characteristics were compared m2 cancer or cervical cancer in situ) as of the date of diagnosis of between cases and controls using tests. SV40 antibody the case. Matching criteria were relaxed in certain cases to status in relation to NHL was assessed in two ways. First, achieve a match: seven controls were up to 2 years older than individuals were defined as positive or negative for SV40 their matched cases, and for four case-control pairs, date of antibodies based solely on the ELISA absorbance cutoff value blood draw differed as much as 1 month. Participants in this of 0.1. To investigate a dose-response relationship, three levels study include 74 NHL cases and 120 controls selected for a of positive absorbance values were defined, based on dividing previous case-control study of pesticide exposure conducted the distribution in the controls into thirds. To account for within these cohorts in 1994 (24). possible immunologic cross-reactivity on the ELISA, a second analysis included results from the blocking assays, where SV40 Laboratory Methods. Virus-like particles (VLP) were puri- antibody-positive samples from the ELISA assay were deter- fied from insect cells infected with recombinant baculoviruses mined to be SV40-specific or cross-reactive, as described expressing the VP1 major capsid protein of SV40, as previously above. In both analyses, the association between SV40 described (25). For ELISA, PolySorp microtiter plates (Nunc, antibody seropositivity and NHL was estimated by calculating Naperville, IL) were sensitized with 20 to 30 ng of VLP protein matched odds ratios (OR) and 95% confidence intervals (95% per well. The serum dilution (1:400) was left to react for 1 hour CI) using conditional logistic regression. ORs were similar at 37jC. Antigen-bound immunoglobulin was detected with using either serum samples collected in 1974 or plasma peroxidase-conjugated antibodies against human immuno- samples collected in 1989; thus, all were combined. Forty- globulin G (Zymed, San Francisco, CA). After 30 minutes at seven of the 170 cases and their matched controls donated room temperature, color development was initiated by the blood to both CLUE I in 1974 and CLUE II in 1989; the variance addition of 2,2V-azino-di-(3-ethylbenzthiazoline-6-sulfonate) of the repeated measures of SV40 antibody levels in this hydrogen peroxide solution. The reaction was stopped after subgroup was incorporated into logistic regression models 20 minutes, and absorbance (A) was measured at 405 nm. using a robust sandwich estimation method (27). Absorbance values of >0.1 were considered positive by ELISA Education, smoking, and body mass index were considered based on previous results (25). The sensitivity and specificity as potential confounders. When these factors were placed of the ELISA for SV40 detection in humans is unknown individually into regression models with SV40 antibody because there are no human reference standards, but the assay serologic status, no appreciable changes in ORs for SV40 demonstrates 100% sensitivity and 100% specificity for SV40 antibodies and NHL were observed; thus, these factors were infection in macaques (25). not included in the final regression models. Analyses were

Cancer Epidemiol Biomarkers Prev 2005;14(6). June 2005 Downloaded from cebp.aacrjournals.org on September 30, 2021. © 2005 American Association for Cancer Research. 1450 SV40 Antibodies and Risk of Non-Hodgkin Lymphoma

stratified by NHL subtype, combining diffuse large B-cell Of the 33 cases and 41 controls who tested positive by ELISA, lymphoma with follicular B-cell lymphoma, after initial 27 (82%) case samples and 31 (76%) control samples exhibited analyses revealed homogenous risk estimates for these sub- cross-reactivity with JCV and/or BKV on the blocking assays types. One case of T-cell lymphoma was identified and com- (Table 2). Two case samples and three control samples bined with the other/unknown group. Analyses were also contained antibodies that were not specific to JCV, BKV, or stratified by two categories of time between blood draw and SV40. Only four case samples and seven control samples diagnosis (<11 and z11 years), defined by the median value in contained SV40-specific antibodies (i.e., antibodies reacted with the cases. All statistical tests were two sided. Analyses were SV40 and did not react with JCV or BKV; Table 2). Whereas the conducted using SAS, version 8 (SAS Institute, Inc., Cary, NC). cross-reactive antibodies were positively associated with NHL (OR, 2.07; 95% CI, 1.03-4.16), there was no statistically significant difference between the proportion of cases and Results controls with antibodies specific to SV40 (OR, 1.51; 95% CI, 0.41-5.52; Table 2). The magnitude of the observed associations Baseline characteristics are presented for NHL cases and was slightly greater when considering only B-cell NHL, matched controls in Table 1. All but one case was White, and although the general pattern of risk was similar (Table 2). there were more females than males, reflecting the race and sex The median time between blood draw and NHL diagnosis distributions in the underlying CLUE cohorts. No statistically was 10.8 years (SD = 7.7 years). Results were similar for cases significant differences between cases and controls were observed diagnosed 11 years before or 11 years after the time of blood for education, smoking status, or body mass index (Table 1). draw (Table 3). Only one case diagnosed within 11 years of SV40 antibody levels in the human sera were low; the blood draw and three cases diagnosed z11 years after blood median absorbance value of the SV40-positive human sera was draw had SV40-specific antibodies, and no statistically 0.23 compared with 1.86 for the macaque sera. Based on the significantly associations were observed with NHL for either dichotomous absorbance cut point of 0.1, SV40 antibodies were of these groups (Table 3). Sixteen cases were diagnosed within detected using the ELISA assay in 15% of cases and 10% of 2 years of blood draw, none of which were positive for SV40- controls (matched OR, 1.97; 95% CI, 1.03-3.76; Table 2). This specific antibodies, and exclusion of these cases did not association was particularly strong for B-cell lymphomas (OR, appreciably change the results (data not shown). Similar 2.85; 95% CI, 1.32-6.17), whereas SV40 antibodies were not results by time to diagnosis were obtained when cases of associated with lymphomas of other/unknown subtypes (OR, other/unknown subtypes were excluded (data not shown). 0.92; 95% CI, 0.30-2.83). There was no dose-response relationship between SV40 antibody levels and overall risk of NHL (Table 2). Discussion

Table 1. Baseline characteristics of NHL cases and matched Individuals positive for SV40 antibodies based on the ELISA controls: Washington County, MD (1975-2002) assay alone had an increased risk for developing NHL, particularly B-cell NHL (OR, 2.85; 95% CI, 1.32-6.17), but Characteristic Cases Controls m2 (P) further investigation showed that the association was due to (n = 170), (n = 340), cross-reactivity with JCV and/or BKV. Presence of antibodies n (%) n (%) specific to SV40, as measured from serum collected years before diagnosis, was not associated with the subsequent Gender* Female 93 (54.7) 186 (54.7) development of NHL. Male 77 (45.3) 154 (45.3) This is the first prospective study of SV40 antibodies and c Age at blood draw (y)*, the development of NHL. In contrast to analyses of NHL Mean (SD) incidence rates by birth cohorts, the measurement of anti- CLUE 1 baseline (1974) 50.9 (12.8) 51.0 (12.8) 0.98 bodies as biomarkers of past SV40 infection provides CLUE 2 baseline (1989) 61.6 (11.4) 61.5 (11.8) 0.91 Race exposure data at the individual level. However, antibodies White 169 (99.4) 337 (99.1) 0.72 themselves are limited in their specificity for SV40 infection Black or other 1 (0.6) 3 (0.9) status. If individuals were exposed to inactivated SV40 b Education through the polio vaccine, they could have produced anti- <12 grades 67 (39.4) 130 (38.5) 0.32 bodies to SV40 capsid proteins in the absence of active 12 grades 62 (36.5) 143 (42.3) infection. In this case, these individuals would be misclassi- >12 grades 41 (24.1) 65 (19.2) Smoking status fied as having past SV40 infection. Measurement of antibodies Never 75 (44.1) 153 (45.0) 0.97 to the T antigen might clarify this issue, because only those Former 55 (32.4) 110 (32.4) with active infection could develop T antibodies. However, Current 40 (23.5) 77 (22.7) 2 x only 4 of 217 NHL case samples and 7 of 434 control samples Body mass index (kg/m ) had antibodies specific to SV40 in this study. Therefore, it is <25 34 (43.0) 58 (36.7) 0.62 25-29 28 (35.4) 60 (38.0) unlikely that a true association between SV40 infection and z30 17 (21.5) 40 (25.3) NHL is masked by nondifferential misclassification of CLUE participation* exposed, noninfected individuals as infected in this small CLUE 1 only (1974) 91 (53.5) 182 (53.5) subset. The small number of participants who had SV40- CLUE 2 only (1989) 32 (18.8) 64 (18.8) specific antibodies limited the interpretation of results CLUE 1 and 2 47 (27.6) 94 (27.6) stratified by NHL subtype. NHL subtype B cell, diffuse large cell 70 (41.2) SV40 antibodies were measured in this study using VLP- B cell, follicular 43 (25.3) based ELISA assays, which were previously validated against T cell 1 (0.6) the plaque neutralization assay for detection of SV40 anti- Other/unknown 56 (32.9) bodies in SV40-infected macaque sera (25). Recent data suggest that the low level of SV40-neutralizing antibodies *Age, gender, and CLUE participation were matching factors. detected in human sera are largely the result of cross- cStudy subjects who participants of both CLUE I and II are counted twice, once for each blood donation. reactivity with BKV and JCV antibodies (28). Therefore, we bInformation on education was missing for two controls. used competitive inhibition or ‘‘blocking’’ assays to deter- xBMI collected in 1989 (CLUE II) only. mine the specificity of SV40 antibodies initially detected by

Table 2. SV40 antibody seroprevalence and risk of incident NHL: Washington County, MD (1975-2003)

Method of measurement: SV40 antibody (Ab) status Cases*, n (%) Controls*, n (%) Matched OR (95% CI)

All NHL: ELISA SV40 Ab-negative 184 (84.8) 393 (90.6) 1.00 SV40 Ab-positive 33 (15.2) 41 (9.5) 1.97 (1.03-3.76) c SV40 Ab-positive, low reactivity 9 (4.1) 13 (3.0) 1.74 (0.67-4.51) SV40 Ab-positive, middle reactivity 17 (7.8) 14 (3.2) 3.30 (1.38-7.91) SV40 Ab-positive, high reactivity 7 (3.2) 14 (3.2) 1.16 (0.43-3.13) ELISA following blocking assays: b SV40-negative Ab 186 (86.9) 396 (91.2) 1.00 Cross-reactive Ab 27 (12.6) 31 (7.1) 2.07 (1.03-4.16) SV40-specific Ab 4 (1.8) 7 (1.6) 1.51 (0.41-5.52) B-cell NHL only: ELISA: SV40 Ab-negative 119 (83.2) 264 (92.3) 1.00 SV40 Ab-positive 24 (16.8) 22 (7.7) 2.85 (1.32-6.17) ELISA following blocking assays: SV40-negative Ab 119 (83.2) 267 (93.4) 1.00 Cross-reactive Ab 21 (14.7) 16 (5.6) 3.50 (1.45-8.46) SV40-specific Ab 3 (2.1) 3 (1.0) 3.22 (0.67-15.54)

Abbreviation: Ab, antibody. *Cases (n = 170) contributed 217 blood samples (47 donated blood in both 1979 and 1989); 340 controls contributed 434 blood samples (94 donated blood in both 1979 and 1989). cThree levels of reactivity were determined based on the distribution of ELISA absorbance values among controls who were positive (absorbance values > 0.1). bIncludes two cases (other/unknown NHL) and three controls (matched to B-cell NHL cases) with samples that tested positive for SV40 antibodies by ELISA but were not inhibited by SV40, JCV, or BKV in blocking assays.

ELISA, an approach used also by other investigators (23, 26). 60 received polio vaccine, with 88% coverage among those Consistent with previous studies, SV40 antibodies detected by 20 years old and younger (29). In the present study, no ELISA in the present study were highly cross-reactive with significant differences were observed in SV40 antibody JCV and BKV VLPs, and the actual prevalence of SV40- seropositivity between those who were younger or older than specific antibodies was low (2%). age 20 in 1961 (data not shown). Human transmission of SV40 Whereas the effect of long-term storage on SV40 antibodies has been proposed as another route of exposure for individ- in human sera has not been documented, there is no reason to uals born after 1963 (30). However, if SV40 infection is truly suspect freezing would differentially affect antibodies to SV40 associated with NHL, then SV40 antibody levels should be versus antibodies to JCV and BKV, which were highly higher among NHL cases than controls in a case-control study, prevalent among study samples. No significant differences in regardless of the route of SV40 infection. SV40 antibody levels were observed between samples collect- Cohort participants who moved out of Maryland would ed in 1974 and 1989. Finally, even if freezing had a small effect have been lost to follow-up, as incident cases were ascertained on antibody levels, no bias would be introduced into the only within the state. However, the population of Washington results, because cases and controls were matched on the County is increasing, and NHL is a relatively rare cancer; thus, number of freeze/thaw cycles of the sera. few cases are likely to have been missed. Additionally, those All study participants were born before 1960; thus, all were who moved out of state had the same opportunity for SV40 potentially exposed to SV40-contaminated polio vaccine. By exposure than those who remained in the state, because all 1961, an estimated 62% of individuals under the age of were potentially vaccinated for polio. Limited information was

Table 3. SV40 antibody seroprevalence and risk of incident NHL by time from blood draw to diagnosis: Washington County, MD (1975-2003)

Method of measurement: SV40 antibody (Ab) status Cases*, n (%) Controls*, n (%) Matched OR (95% CI)

Blood draw <11 years before diagnosis: ELISA: SV40 Ab-negative 93 (83.0) 203 (91.4) 1.00 SV40 Ab-positive 18 (16.2) 19 (8.6) 2.28 (1.08-4.81) ELISA following blocking assays: c SV40-negative Ab 95 (85.6) 203 (91.4) 1.00 Cross-reactive Ab 15 (13.5) 15 (6.8) 2.35 (1.03-5.37) SV40-specific Ab 1 (0.9) 4 (1.8) 0.55 (0.06-4.80) Blood draw z11 y before diagnosis: ELISA: SV40 Ab-negative 91 (85.9) 190 (89.6) 1.00 SV40 Ab-positive 15 (14.2) 22 (10.4) 1.63 (0.66-4.03) ELISA following blocking assays: c SV40-negative Ab 91 (85.8) 193 (91.0) 1.00 Cross-reactive Ab 12 (11.3) 16 (7.6) 1.99 (0.75-5.28) SV40-specific Ab 3 (2.8) 3 (1.4) 3.08 (0.49-19.31)

Abbreviation: Ab, antibody. *Cases (n = 170) contributed 217 blood samples (47 donated blood in both 1979 and 1989); 340 controls contributed 434 blood samples (94 donated blood in both 1979 and 1989). cIncludes two cases (<11 years prior to diagnosis) and three controls (z11 years prior to diagnosis) with samples that tested positive for SV40 antibodies by ELISA but were not inhibited by SV40, JCV, or BKV in blocking assays.

Cancer Epidemiol Biomarkers Prev 2005;14(6). June 2005 Downloaded from cebp.aacrjournals.org on September 30, 2021. © 2005 American Association for Cancer Research. 1452 SV40 Antibodies and Risk of Non-Hodgkin Lymphoma

available for other suspected risk factors for NHL at baseline, 10. Martini F, Dolcetti R, Gloghini A, et al. Simian-virus-40 footprints in human lymphoproliferative disorders of HIV and HIV+ patients. Int J Cancer including birth order, income, and history of infections. These 1998;78:669 – 74. factors, if associated with polio vaccination, could have been 11. Nakatsuka S, Liu A, Dong Z, et al. Simian virus 40 sequences in malignant negative confounders of the association between SV40 infec- lymphomas in Japan. Cancer Res 2003;63:7606 – 8. tion and NHL, assuming that polio vaccination was the main 12. Shivapurkar N, Harada K, Reddy J, et al. Presence of simian virus 40 DNA source of SV40 exposure. sequences in human lymphomas. Lancet 2002;359:851 – 2. 13. Vilchez RA, Madden CR, Kozinetz CA, et al. Association between simian Our results indicate that past SV40 infection is not virus 40 and non-Hodgkin lymphoma. Lancet 2002;359:817 – 23. associated with the development of NHL. The observed 14. Rizzo P, Carbone M, Fisher SG, et al. Simian virus 40 is present in most association between NHL and antibodies to SV40 that cross- United States human mesotheliomas, but it is rarely present in non- react with JCV and/or BKV suggests that these polyomavi- Hodgkin’s lymphoma. Chest 1999;116:470 – 3S. 15. Capello D, Rossi D, Gaudino G, Carbone A, Gaidano G. Simian virus 40 ruses should be examined further as potential risk factors for infection in lymphoproliferative disorders. Lancet 2003;361:88 – 9. NHL. Future epidemiologic studies of polyomavirus anti- 16. MacKenzie J, Wilson KS, Perry J, Gallagher A, Jarrett RF. Association bodies and cancer should also incorporate biomarkers of between simian virus 40 DNA and lymphoma in the United kingdom. J Natl immunologic cross-reactivity with other oncogenic viruses to Cancer Inst 2003;95:1001 – 3. 17. Yazdi AS, Puchta U, Flaig MJ, Sander CA. Lack of evidence for the presence investigate the potential interplay between viruses in the of Simian virus 40 DNA in cutaneous lymphomas. J Invest Dermatol 2003; etiology of NHL. 121:212 – 3. 18. Engels EA, Rodman LH, Frisch M, Goedert JJ, Biggar RJ. Childhood exposure to simian virus 40-contaminated poliovirus vaccine and risk of AIDS-associated non-Hodgkin’s lymphoma. Int J Cancer 2003;106:283 – 7. Acknowledgments 19. Engels EA, Katki HA, Nielsen NM, et al. Cancer incidence in Denmark We thank Judith Hoffman-Bolton at the George W. Comstock Center following exposure to poliovirus vaccine contaminated with simian virus 40. for Public Health Research and Prevention for assistance with J Natl Cancer Inst 2003;95:532 – 9. participant selection; the Clinical Immunization Safety Assessment 20. Holly EA, Bracci PM. Population-based study of non-Hodgkin lympho- Network for technical and administrative support, especially Christine ma, histology, and medical history among human immunodeficiency Casey and Philip LaRussa for reviewing a draft of the article; the virus-negative participants in San Francisco. Am J Epidemiol 2003;158: 316 – 27. laboratory assistance of Barbara Clayman (Johns Hopkins University 21. Rollison DE, Page WF, Crawford H, et al. Case-control study of cancer School of Medicine, Baltimore, MD); and all who participated in the among US Army veterans exposed to simian virus 40-contaminated CLUE cohorts. adenovirus vaccine. Am J Epidemiol 2004;160:317 – 24. 22. de Sanjose S, Shah KV, Domingo-Domenech E, et al. Lack of serological evidence for an association between simian virus 40 and lymphoma. Int J Cancer 2003;104:522 – 4. References 23. Engels EA, Viscidi RP, Galloway DA, et al. Case-control study of simian 1. American Cancer Society. Cancer facts and figures 2004. Atlanta (GA): virus 40 and non-Hodgkin lymphoma in the United States. J Natl Cancer American Cancer Society. Ref Type: Report; 2004. Inst 2004;96:1368 – 74. 2. Dal Maso L, Franceschi S. Epidemiology of non-Hodgkin lymphomas and 24. Rothman N, Cantor KP, Blair A, et al. A nested case-control study of non- other haemolymphopoietic neoplasms in people with AIDS. Lancet Oncol Hodgkin lymphoma and serum organochlorine residues. Lancet 1997;350: 2003;4:110 – 9. 240 – 4. 3. Gisbert JP, Garcia-Buey L, Pajares JM, Moreno-Otero R. Prevalence of 25. Viscidi RP, Rollison DE, Viscidi E, et al. Serological cross-reactivities hepatitis C virus infection in B-cell non-Hodgkin’s lymphoma: systematic between antibodies to simian virus 40, BK virus, and JC virus assessed by review and meta-analysis. Gastroenterology 2003;125:1723 – 32. virus-like-particle-based enzyme immunoassays. Clin Diagn Lab Immunol 4. Young LS, Murray PG. Epstein-Barr virus and oncogenesis: from latent 2003;10:278 – 85. genes to tumours. Oncogene 2003;22:5108 – 21. 26. Carter JJ, Madeleine MM, Wipf GC, et al. Lack of serologic evidence for 5. Vilchez RA, Butel JS. SV40 in human brain cancers and non-Hodgkin’s prevalent simian virus 40 infection in humans. J Natl Cancer Inst 2003;95: lymphoma. Oncogene 2003;22:5164 – 72. 1522 – 30. 6. Shah K, Nathanson N. Human exposure to SV40: review and comment. Am 27. Wei LJ, Lin Dy, Weissfeld L. Regression analysis of multivariate incomplete J Epidemiol 1976;103:1 – 12. failure time data by modeling marginal distributions. J Am Stat Assoc 1989; 7. Shah KV. Polyoma viruses (JC virus, BK virus, and simian virus 40) and 84:1065 – 78. human cancer. In: Goedert JJ editor. Infectious causes of cancer: targets for 28. Shah KV, Galloway DA, Knowles WA, Viscidi RP. Simian virus 40 (SV40) intervention. Totowa (NJ): Humana Press, Inc.; 2001. p. 461 – 74. and human cancer: a review of the serological data. Rev Med Virol 2004; 8. Diamandopoulos GT. Leukemia, lymphoma, and osteosarcoma induced in 14:231 – 9. the Syrian golden hamster by simian virus 40. Science 1972;176:173 – 5. 29. Shah K, Nathanson N. Human exposure to SV40: review and comment. Am 9. Daibata M, Nemoto Y, Kamioka M, Imai S, Taguchi H. Simian virus 40 in J Epidemiol 1976;103:1 – 12. Japanese patients with lymphoproliferative disorders. Br J Haematol 2003; 30. Butel JS, Jafar S, Wong C, et al. Evidence of SV40 infections in hospitalized 121:190 – 1. children. Hum Pathol 1999;30:1496 – 502.

Cancer Epidemiol Biomarkers Prev 2005;14(6). June 2005 Downloaded from cebp.aacrjournals.org on September 30, 2021. © 2005 American Association for Cancer Research. Markers of Past Infection with Simian Virus 40 (SV40) and Risk of Incident Non-Hodgkin Lymphoma in a Maryland Cohort

Dana E. Rollison, Kathy J. Helzlsouer, Neal A. Halsey, et al.

Cancer Epidemiol Biomarkers Prev 2005;14:1448-1452.

Updated version Access the most recent version of this article at: http://cebp.aacrjournals.org/content/14/6/1448

Cited articles This article cites 26 articles, 3 of which you can access for free at: http://cebp.aacrjournals.org/content/14/6/1448.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://cebp.aacrjournals.org/content/14/6/1448.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cebp.aacrjournals.org/content/14/6/1448. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.