The inadvertent and unrecognized presence of simian virus 40 (SV40) in the commercial inactivated Salk polio vaccines administered between 1955 and 1963 resulted in the potential exposure of millions of individuals, in the United States and elsewhere, to this polyoma virus of the rhesus macaque [1]. Many of the monkeykidney cultures used to prepare poliovirus pools were infected with the indigenous SV40. Soon after its discovery in 1960 [2],SV40 was found to be oncogenic in laboratory animals [3]. Therefore, the possibility that SV40 may cause human disease,particularly cancers, has been a topic of interest since the 1960s [4]. This debate has intensified during the past decade because several groups of investigators, using polymerase chain reaction (PCR) amplification methodology, have detected SV40 genomic sequences in a number of human cancers. These investigators have suggested that the virus contributes to the development of mesothelioma,  osteosarcoma, pediatric and adult brain tumors,and non-Hodgkin lymphomas . The reported presence of SV40 in tumors in individuals born after 1963 would seem to imply that SV40 is now established as a human infection circulating in communities via person-to-person contact .Other investigators have been skeptical of these claims ; several groups have not been able to detect SV40 sequences in the aforementioned tumors  and epidemiologic studies have not revealed an increased risk of these cancers in populations exposed to SV40-contaminated polio vaccines or adenovirus vaccines . In addition to a large number of scientific publications, the controversy has spawned a report by the Institute of Medicine  a book in the popular press and litigations. The evidence for the pathogenicity of SV40 in humans can be conveniently examined in 3 parts: (1) the nature of the human response after exposure to SV40, (2) the evidence that infection with SV40 has become established in humans, and (3) the evidence that infection with SV40 contributes to the development of human cancers

                             That individuals exposed to SV40 may develop a transient infection was documented during the 1960s. Morris et al found that, after intranasal inoculation of live SV40 (which was an inadvertent contaminant of an experimental respiratory syncytial virus vaccine), ∼60% of the volunteers developed neutralizing antibodies to the virus. The antibody titers in the positive sera were low: small amounts of infectious virus were recovered from throat swabs of some of the individuals 7 or 11 days after inoculation; virus was not recovered from rectal swabs.After oral administration of SV40 (which was a contaminant of the experimental lots of the Sabin attenuated polio vaccine) to children, small amounts of virus were recovered intermittently, for up to 5 weeks, from the stools of some of the volunteers, but none of the children developed an antibody response .Both the serological study of American zoo workers that is reported by Engels et al. in this issue of the Journal and an earlier study of employees of monkey export firms in India  suggest that infection with SV40 may also be acquired by contact with (presumably) naturally infected primates. Engels et al.  were able to show (1) that the prevalence of SV40 antibody in “nonhuman-primate zoo workers” (i.e., “those currently working specifically with either nonhuman primates or a larger class of animals including nonhuman primates and those in senior administrative positions, which were assumed to be filled by individuals with extensive animal-handling experience”) was higher than that in “other zoo workers” (i.e., “those currently working with classes of animals not including nonhuman primates and those performing maintenance, clerical, or visitor service”) and (2) that the SV40-reactive antibodies in the sera of the non human primate zoo workers were more likely to be SV40 specific (see below) than were those in the sera of the other zoo workers.

                             Although SV40 may produce transient infection in exposed individuals, the results of serological and virological studies indicate that SV40 has not become established in the human population. It is very likely that the low levels of SV40-reactive antibodies described in human sera in many previous investigations  are the result of cross-reactivity between SV40 and the widely prevalent human polyomaviruses BKV and JCV . Sera of rhesus macaques naturally infected with SV40 react strongly with SV40 viruslike particles (VLPs), but they also react—to a lesser degree but unambiguously—with BKV VLPs and with JCV VLPs [35, 36]. Rhesus serum’s reactivity with BKV VLPs and with JCV VLPs is decreased by preadsorption with both SV40 VLPs and either BKV VLPs or JCV VLPs. Conversely, the low-level SV40 reactivity of human sera was clearly correlated with reactivity to BKV and JCV and, whenever measured, was significantly decreased by preadsorption with either BKV VLPs or JCV VLPs .

                             The most puzzling aspect of the controversy —and the heart of the problem—has been a lack of agreement on whether authentic SV40 sequences are present in  human tumors. The studies with positive results have reported small copy numbers(often estimated as being !1 copy/10–100+ cells) of T antigen–coding sequences in a wide variety of unrelated tumors. Relatively few studies have attempted to identify SV40 transcripts or T protein in the tumor tissue, and the results of these studies have been inconclusive. On the other hand, employing similar or more-sensitive methods, several recent studies of mesothelioma, lymphomas, and brain tumors have shown largely or completely negative results. Contamination with laboratory plasmids has now been identified as one reason for the discrepancy.At a 1997 meeting organized by the US Food and Drug Administration to examine SV40 as a possible human pathogen, Griffiths et al. and Volter et al.  independently suggested the possibility that false-positive results may follow such contamination, because fragments of the SV40 genome—especially of the early region coding for large T antigen—have been used in the construction of hundreds of expression vectors worldwide.