Lack of serologic evidence for prevalent simian virus 40 infection in humans

Background: Propagation of poliovirus in monkey kidney cells led to the inadvertent contamination of poliovirus vaccines with simian virus 40 (SV40) between 1955 and 1963. Recent studies using polymerase chain reaction-based strategies have detected SV40 DNA in a large number of tumor types. The finding of SV40 DNA in tumors from individuals who are too young to have been exposed to SV40-contaminated vaccines has led to the suggestion that SV40 has become a prevalent transmissible human pathogen. To test this hypothesis, we screened human sera for antibodies to SV40 using direct and competitive enzyme-linked immunosorbent assays (ELISAs). Methods: An ELISA was developed using recombinant SV40 virus-like particles (VLPs) and was validated using sera from naturally infected macaques. VLPs of SV40 and the related ubiquitous human polyomaviruses, JCV and BKV, were used to screen human sera to determine the prevalence of SV40, JCV, and BKV antibodies among a normal population of control subjects (n = 487) and among case patients with either osteosarcoma (n = 122) or prostate cancer (n = 90). A competitive ELISA in which sera were pre-adsorbed with each type of VLP was used to identify cross-reactive antibodies. Correlations of reactivity among the three polyomavirus types were calculated using the Spearman correlation coefficient. All statistical tests were two-sided. Results: BKV and JCV antibodies were prevalent in all case patients and control subjects examined. In contrast, only 6.6% (46/699) of serum samples were positive for SV40 antibodies by ELISA; however, none of these samples could be confirmed as having authentic SV40 antibodies following pre-adsorption with JCV or BKV VLPs. Conclusion: These data indicate that some individuals have BKV and/or JCV antibodies that cross-react with SV40, but they do not provide support for SV40 being a prevalent human pathogen.