In earlier postings, we recounted how in 1959 Bernice Eddy, at the U. S. National Institutes of Health (NIH), and then Maurice Hilleman, at Merck & Co, discovered a new virus, simian virus 40 (SV40), in early lots of the Salk and Sabin polio vaccines (1, 2). The virus inadvertently contaminated those vaccines because it was unknowingly present in the rhesus monkey kidney cell cultures in which the vaccines were grown. Hilleman gave SV40 its name. It was the 40th simian virus that the Merck lab found in its rhesus kidney cell cultures.
Next, in 1961, both Eddy and Hilleman discovered that inoculating SV40 into hamsters caused cancer in about half of the animals. But, by then, hundreds of millions of people worldwide had been inoculated with live SV40, via the contaminated polio vaccines!
[Aside 1: Since the Sabin vaccine contains live, attenuated poliovirus, whereas the Salk vaccine contained formaldehyde-inactivated poliovirus, it was initially thought that any issues stemming from SV40 contamination would be limited to the Sabin vaccine. However, about one in 10,000 SV40 particles survived the formaldehyde treatment used to inactivate poliovirus in the Salk vaccine. Thus, early lots of both the Salk vaccine and the Sabin vaccine were contaminated with live SV40. But, despite the initial belief that that only the Sabin vaccine posed a threat of SV40 contamination, recipients of the Salk vaccine, but not those receiving the Sabin vaccine, developed an antibody response against SV40. Apparently, any SV40 that may have been present in the orally administered Sabin vaccine was killed in the intestinal tracts of its recipients. But, while the antibody response seen in recipients of the injection-administered Salk vaccine would be consistent with an actual SV40 infection, that response might have been only against the SV40 in the vaccine itself (see the main text).]
In July 1961, the New York Times broke part of this story, reporting that Merck was withdrawing its vaccines because they were contaminated with a monkey virus. However, the Times article did not mention the cancer connection. In fact, the Times did not report that aspect of the story until more than a year later, causing some individuals to suspect that the government deliberately withheld that information from the public.
I do not know whether or not there was a deliberate intent by the government to withhold information. Joseph Smadel, Bernice Eddy’s immediate superior at the NIH and himself an eminent scientist, dismissed Eddy’s finding of a tumor-inducing virus in early lots of the Salk vaccine, after he reviewed her data. Later, after Eddy reported her findings at an October 1960 cancer conference in New York, Smadel forbade her from speaking about the matter again in public without first clearing her remarks with him. However, Smadel was indeed concerned when similar findings were reported to him later by Hilleman. Was sexism a factor in the dismissal of Eddy’s results? Perhaps, but it also was alleged (in fact by Hilleman) that Eddy’s experiments were poorly controlled.
At any rate, Eddy had, in fact, presented her findings at an open scientific conference. What’s more, at around the same time, Hilleman too presented his findings in public, at a conference in Copenhagen. So, their discovery, while not purposely publicized, was not kept secret either.
Next, consider that the second report in the New York Times, which mentioned the cancer connection for the first time, was buried on page 27 of the newspaper. This fact points more to the inability of the press to appreciate and cover a complex scientific issue, than to an attempt by the government to suppress the story.
In any event, the government did not alert the public to the possible danger that might be lurking in the polio vaccines. Is it possible that the NIH did not appreciate that threat? This seemingly implausible explanation is consistent with the fact that the NIH did not begin to screen all new lots of the polio vaccines for SV40 until 1963.
Another perhaps more likely possibility is that the government simply believed that alerting the public might have irrevocably broken its confidence in the vaccines, ultimately causing far more disease than might have been caused by the vaccines themselves. Even so, an impending public health debacle, of unprecedented severity, could not yet have been ruled out.
We return to our main story after a bit of background on SV40.
SV40 is a member of the polyomavirus family of small, double-stranded DNA viruses. It is one of several polyomaviruses that can transform normal cells into tumor cells in cell culture, as well as induce tumors in laboratory animals. For that reason, and because SV40 is a relatively simple virus (its genome contains only about 5,200 bases pairs), it was intensively studied for what it might reveal about tumor genesis. Moreover, it also came to serve as an important model to investigate fundamental issues in eukaryotic molecular biology (3).
The polyomaviruses are widespread in their natural hosts, in which they give rise to lifelong, usually benign, persistent infections. The Asian rhesus macaque is the natural host for SV40. These facts explain why SV40 was not evident in the rhesus monkey kidney cell cultures that were used to propagate the early poliovirus vaccine lots. SV40 does not cause sufficient cytopathology in those rhesus macaque cell cultures to reveal its presence in them. However, culture fluids from those rhesus cell cultures caused extensive cytopathology when added to African green monkey kidney cell cultures. Indeed, that is how SV40 was discovered.
[Aside 2: The Human Polyomaviruses The JC polyomavirus (JCPyV) and the BK polyomavirus (BKPyV), each discovered in 1971, are the best known polyomaviruses that naturally infect humans. JCPyV and BKPyV are each ubiquitous in their human host, in which they typically give rise to lifelong, benign, persistent infections. Yet, JCPyV can give rise to a rare but fatal demyelinating disease, progressive multifocal encephalopathy (PML), in immunologically compromised individuals. And BKPyV can cause kidney disease, also in immunologically compromised persons.
More recently, several additional human polyomaviruses have been discovered, by means of modern DNA amplification procedures. One of these viruses, the ubiquitous Merkel cell polyomavirus (MCPyV), is associated with a rare, aggressive human malignancy, Merkel cell carcinoma, and is the best candidate for an oncogenic human polyomavirus.]
We now resume our main story, with the following key points.
Despite the fact that the unintended exposure of millions of individuals to SV40 via the contaminated polio vaccines in the 1950s posed a potential public health crisis of immense proportions, it still is not clear whether SV40 is an agent of human disease. Moreover, it is not known whether SV40 is circulating in the human population. How can this be?
Early investigations into this matter in the 1960s were compromised by the fact that it was not apparent which individuals had actually received SV40-contaminated vaccines and which did not. That was so in part because the serological reagents and procedures of the day were not sensitive or accurate enough to generate unambiguous results. Additionally, population sample sizes were often too small to generate statistically significant results. [This was especially so in the case of the rare childhood tumors in which SV40 had been implicated.] And, since cancer is a disease that may take decades to emerge, it was possible that more time needed to pass before the virus might unequivocally reveal itself as a cause of human cancer. At any rate, since the substantial experimental data then available could neither establish nor absolve SV40 as a cause of cancer in humans, the NIH conceded that more research and better methods for detecting the virus would be needed to settle the issue.
The more recent development of extremely sensitive polymerase chain reaction (PCR)-based procedures, which can detect minute levels of specific DNA sequences, led to renewed interest in whether SV40 might be present in humans, and whether it might be an agent of human disease. Using PCR technology, several different research groups detected SV40 DNA in four types of human cancers; mesotheliomas, osteosarcomas, non-Hodgkin’s lymphomas, and childhood brain tumors. These findings were alarming because the four tumor types, in which SV40 was detected in humans, are the same tumors that SV40 induces experimentally in hamsters (i.e., mesothelioma, bone, lymphoma, and brain).
PCR procedures also detected SV40 DNA in individuals who never were inoculated with an SV40-contaminated vaccine. This too was disturbing because it raised the specter that SV40 might be circulating in the human population, spreading by horizontal transmission from one individual to another.
Yet the issue of SV40 in humans remains controversial because other studies, from other research groups, using similar PCR procedures, could not detect SV40 DNA in human tissues. In addition, newer, more sensitive and accurate serologic procedures could not demonstrate to everyone’s satisfaction that SV40 circulates in humans.
How might we explain how capable scientists, using powerful and proven techniques, can obtain such disparate experimental results? Ironically, the sensitivity of PCR itself may be a problem, since it increases the likelihood of false-positive results, which may occur from the slightest sample contamination. Thus, it is important to have suitable positive and negative control samples that might be processed side-by-side with test samples; a sometimes difficult criterion to fulfill. [Bearing the above in mind, consider the following example, in which a human mesothelioma sample was micro-dissected to separate normal tissue from the actual tumor. SV40 sequences were detected in the tumor, but not in the adjacent normal tissue, which served as an internal control.]
Another potential source of error stems from the widespread prevalence of human polyomaviruses (e.g., JCPyV, BKPyV, and MCPyV) in the human population, leaving open the possibility that these viruses, rather than SV40, are detected by the PCR-based procedures. But, with that possibility in mind, several researchers took the extra step of confirming the presence of SV40 sequences by direct sequencing of the PCR-amplified DNA. The ubiquitous human polyomaviruses are also a concern when carrying out serological procedures, since immune cross-reactivity between SV40 and these viruses remains a potential source of error.
Another problem is theoretical rather than technical. An underlying premise behind these studies is that the continued presence and expression of polyomaviral tumor genes are necessary for a tumor cell to express its tumor cell characteristics. Indeed, this was shown to be the case fifty years ago for cells transformed in culture by polyomaviruses. Thus, the absence of SV40 DNA, or SV40 tumor antigens, in a tumor is taken as evidence against SV40 as the cause of the tumor. However, there is some experimental evidence that the paradigm itself may not always be true. Cancers result from a complex multistage course of events and, in some instances, the virus may play a necessary role only at a particular point in the overall process. Thus, the absence of SV40 DNA, or SV40 antigens, in a tumor may not be definitive proof against viral involvement in the tumor process.
The above points help us to appreciate why there is no consensus regarding a role for SV40 in human cancer, and indeed whether SV40 might be circulating in humans. Yet we remain troubled by the fact that SV40 can transform a variety of cells in culture and can induce tumors in laboratory animals. And there are additional experimental findings that while contentious, cannot be easily ignored. For instance, the types of human cancers, in which SV40 DNA was detected by some researchers, are the same types of tumors that SV40 induces in laboratory animals. Also, infectious SV40 was isolated from a brain cancer of a 4-year-old child. [LT/pRb and LT/p53 complexes were identified in human brain tumors, consistent with current understanding of how SV40 induces neoplasia (3).]
Yet, notwithstanding the force of the above arguments, impressive evidence has been presented against a role for SV40 in human cancer. And, if SV40 indeed were responsible for human cancer on a large scale, then it is rather certain that there would be little if any uncertainty in that regard.
1. Jonas Salk and Albert Sabin: One of the Great Rivalries of Medical Science, posted on the blog, March 27, 2014
2. Maurice Hilleman: Unsung Giant of Vaccinology, posted on the blog, April 24, 2014
3. Virology: Molecular Biology and Pathogenesis, Leonard C. Norkin, ASM Press, 2010.