Studies Find No Evidence That SV40 is Related to Human Cancer
Recent studies by scientists at the National Cancer Institute (NCI), one of the National Institutes of Health, provide further evidence that exposure to simian virus 40 (SV40) is not associated with cancer in humans. Many U.S. polio vaccines administered from 1955 to1962 were accidentally contaminated with SV40 because the vaccines were grown in monkey kidney tissue. Before discovery of the virus led to changes in vaccine manufacture, millions of Americans received SV40-contaminated polio vaccines. This has been a significant public health concern, as SV40 has been shown to cause cancer in experimental animals.
Studies investigating the possible connection between SV40 and human cancer have been inconclusive. For example, some laboratories have reported the detection of minute quantities of SV40 DNA in human tumors while others have not been able to replicate these results.
One approach used by NCI scientists relies on detection of antibodies. In reaction to exposure to a foreign particle such as a virus, the body creates antibodies, which are proteins that remain in the blood long after exposure and may help the body fight back against the virus in the future. Because antibodies remain in the body for so long, scientists can test for their presence to determine whether someone has been exposed to a particular virus, even decades earlier.
In the first study1 investigators led by Eric Engels, M.D., in NCI's Division of Cancer Epidemiology and Genetics, examined the possible association between SV40 exposure and non-Hodgkin lymphoma. In laboratory rodents, SV40 causes lymphoma, a cancer of the lymph nodes. Because some laboratory studies report that SV40 DNA can be detected in tumors from humans with this cancer, NCI scientists tested for SV40 antibodies in the blood of 724 non-Hodgkin lymphoma patients and in 622 controls--people without the cancer. Overall, SV40-reactive antibodies were found in only 7-10 percent of cancer cases and 10-11 percent of controls, indicating no statistical association between the virus and non-Hodgkin lymphoma. These results were published in the September 15, 2004 Journal of the National Cancer Institute.
Some antibodies are very specific (reacting against, for example, only one kind of virus) while others show more general activity (reacting against many viruses). Many older SV40 studies have used antibody tests that do not distinguish between antibodies specific to SV40 and other, similar antibodies, such as those reacting to some common human viruses. For this study, Engels and his group used a new, much more sensitive and specific SV40 antibody test. Called the virus-like particle assay, or VLP, this test uses particles structurally similar to SV40 to detect the presence of antibodies specific to the SV40 capsid (outer coat). They also used VLPs for other viruses similar to SV40 to block any generally reactive SV40 antibodies. After blocking, they still found no difference in the prevalence of antibodies to SV40 between lymphoma cases and controls.
While the NCI scientists were not able to compare the presence of antibodies in their results directly with the presence of actual SV40 DNA in subjects' tumor tissue, Engels believes the study is strong. Two laboratories, whose personnel did not know whether the samples they tested were from subjects with cancer or not, produced similar results. "Future studies should compare the results of antibody tests directly with the results of tests for SV40 DNA in tumor tissue," Engels said. "Additional tests for antibodies to other SV40 proteins would be a valuable tool for this research area."
More recently, in a paper published in the February 2005 issue of Cancer, Epidemiology, Biomarkers, and Prevention 2, Engels and his colleagues followed up on this research using another antibody assay, which detects antibodies against T antigen, an SV40 protein directly implicated in cellular transformation. Animals with SV40-induced tumors often make high levels of antibody against T antigen. The study included 85 people with NHL and 95 controls chosen from the larger U.S.-based case-control study. NCI researchers found only low levels of antibodies to T antigen in 5 people with NHL and 5 controls. Therefore, there was no association found between the presence of antibodies to T antigen and NHL in humans.
In another study3 , published in the August 2004 American Journal of Epidemiology and led by Engels, the researchers also used the VLP test. Because some laboratory studies report that SV40 DNA can be detected in various childhood tumors, NCI scientists evaluated cancer risk of 54,796 United States children whose mothers received polio vaccines during pregnancy before 1963. The mothers may have become infected with SV40 from the vaccines and may have transmitted the virus to their children in the womb or soon after birth. The scientists postulated whether early-life transmission of infection from a mother to her child might be related to the later development of childhood cancer.
Engels and his colleagues compared cancer risk in children whose parents received pre-1963 polio vaccine with cancer risk in children whose mothers did not receive vaccine. They also measured SV40 antibodies in the mothers of 50 of these children who developed cancer and the mothers of 200 children without cancer. One of this study's strengths is the researchers' use of two different means of detecting SV40 antibodies -- the highly specific VLP assay and an older method, long considered the gold standard, called a plaque neutralization assay.
The investigators found that the incidence of cancers of the nervous system and the blood (mainly leukemia), was roughly 2.5 times higher in children whose mothers received pre-1963 vaccine than in children whose mothers did not. However, the pattern of cancers in children whose mothers received the vaccine was not what would be predicted if SV40 caused cancer -- the types of cancers varied and did not correspond to the types in which SV40 has been detected. Also, few women had antibodies to SV40 by either of the antibody tests, and there was no consistent relationship between the development of SV40 antibodies during pregnancy and cancer in children.
"If SV40 were the cause of the tumors in these children, we would expect to see more cases of the types of cancer hypothesized to be linked to SV40 -- for example ependymomas, choroid plexus tumors, and osteosarcomas," Engels explained. "It was notable that among our subjects, we observed only one brain tumor -- an ependymoma -- of the kind in which SV40 DNA has reportedly been detected, and that was in a child whose mother had not received pre-1963 polio vaccine."
The scientists acknowledged that their study was limited by the small number of subjects with cancer and the small number of individual cancer types. Nonetheless, Engels summarized, "Overall, these results argue against an important role for SV40 in childhood cancers."
Engels and scientists at other institutions also evaluated cancer risk in veterans exposed to SV40 in an early U.S. Army adenovirus vaccine given between 1959 and 1961. This study4 , also published in the August 2004 American Journal of Epidemiology,, is the first follow-up study of recipients of SV40-contaminated adenovirus vaccine. The adenovirus vaccine was administered to new military recruits to prevent epidemics of respiratory disease caused by adenovirus arising from crowded living conditions during basic training. Like the polio vaccine, adenovirus vaccine was grown in monkey kidney tissue. However, because adenovirus cannot grow in such tissue without SV40 acting as a helper virus, almost all batches of this vaccine probably contained SV40. The adenovirus vaccine was given to Army recruits during two well-defined time-periods, alternating with three periods of non-use, during the period 1959 to 1961. This alternating pattern allowed for a rigorous comparison of cancer risk in vaccine-exposed and unexposed servicemen in this study.
The researchers used Veterans Administration and military records to identify individuals with cancer and to classify them with respect to receipt of the Army's adenovirus vaccine. The study included cases of mesothelioma, brain tumors, and non-Hodgkin lymphoma, which are tumor types previously hypothesized to be linked to SV40. Importantly, the investigators did not find evidence that SV40-contaminated adenovirus vaccine was associated with an increased risk for these cancers.
The authors noted some limitations of the study. There were only 10 mesothelioma cases and the numbers of cases for individual types of brain tumors were small. These small numbers limit the study's statistical power. Nonetheless, the numbers for all types of brain tumors combined, and for non-Hodgkin lymphoma, were substantial. Additionally, most subjects would also have received polio vaccine possibly contaminated with SV40. However, not every dose of polio vaccine contained SV40, and exposure to polio vaccine was not dependent on whether individuals received adenovirus vaccine. Thus, while the widespread use of polio vaccine could have diluted an association between adenovirus vaccine and cancer, this effect was likely minor. Engels concluded, "Our results should be reassuring to military veterans of the 1950s and 1960s, some of whom received adenovirus vaccines. This study did not find that their exposure to this vaccine was related to an increased risk of cancer."
Finally, in related work 5, Engels and others found that of 254 zoo workers, those exposed to nonhuman primates at work had a higher prevalence of SV40 antibodies than the others: ten percent of those exposed had antibodies specific to SV40, while only three percent of the others did. Engels said, "Most people would not acquire SV40 through this route--in most parts of the world, including the U.S., exposure to nonhuman primates infected with SV40 is rare." He notes that he and his colleagues did not have information about how zoo workers might have been exposed to the virus, for example through scratches, bites, or splashes, and that they could not determine whether these antibodies represent ongoing SV40 infection. Nonetheless, the study suggests that individuals who have exposures to SV40-infected nonhuman primates may represent a useful population in which to study possible health effects of SV40 infection.
In conclusion, although SV40 causes cancer in laboratory animals, substantial epidemiological evidence has accumulated to indicate that SV40 likely does not cause cancer in humans. However, additional laboratory research is needed to better define methods for SV40 detection, as laboratory studies looking for SV40 DNA in human tumors have offered conflicting results. There is also a need to conduct additional studies evaluating cancer patients and controls for antibodies to SV40, which would be present in cancer patients if SV40 causes cancer.
For more information about cancer, please visit the NCI Web site at http://www.cancer.gov or call NCI's Cancer Information Service at 1-800-4-CANCER (1-800-422-6237).
1Engels EA, Viscidi RP, et al. "Case-Control Study of Simian Virus 40 and Non-Hodgkin Lymphoma in the United States." Journal of the National Cancer Institute. Vol. 96, No. 18. September 15, 2004.
2Engels EA, Chen J, et al. "Antibody Responses to Simian Virus 40 T antigen: A Case-Control Study of Non-Hodgkin Lymphoma." Cancer, Epidemiology, Biomarkers, and Prevention, February 2005.
3Engels EA, Chen J al. "Poliovirus Vaccination during Pregnancy, Maternal Seroconversion to Simian Virus 40, and Risk of Childhood Cancer." American Journal of Epidemiology. Vol. 160, No. 4. August 2004.
4Rollison DEM, Page WF, et al. "Case-Control Study of Cancer among US Army Veterans Exposed to Simian Virus 40-contaminated Adenovirus Vaccine." American Journal of Epidemiology. Vol. 160, No. 3. August 2004.
5Engels EA, Viscidi RP, et al. "Serological Evidence for Exposure to Simian Virus 40 in North American Zoo Workers." Journal of Infectious Diseases. December 15, 2004.