Report from the AIDS/Poliovirus Advisory Committee

We have been asked by the Wistar Institute to evaluate the probability that the hypothesis raised by Tom Curtis in an article published in the Rolling Stone magazine of March 19, 1992 under the heading, "The origin of AIDS" may be correct. Briefly stated, the article suggests that the crucial event in the origin of AIDS epidemic was the extensive clinical trial of a live, attenuated poliovirus vaccine conducted by Dr. H. Koprowski and colleagues in Congo in 1957. Since this vaccine was grown in primary monkey kidney cell cultures, and there is evidence to suggest that human immunodeficiency virus (HIV, the etiological agent of AIDS) may have originated from a simian immunodeficiency virus (SIV) ancestor, Mr. Curtis suggested that contamination of these monkey kidney cultures with SIV or an SIV-like virus could have led to the inadvertent inoculation of this virus, together with the attenuated poliovirus, into the mouths of the vaccinated individuals. Subsequent evolution of this virus in infected individuals could have led to the emergence and spread of the currently predominant HIV strains. The timing of these clinical trials, and the general belief that HIV appeared first in central Africa, lend superficial support to this hypothesis.

In evaluating the credibility of this hypothesis, we have taken the approach of assessing the probability that each step in this postulated mode of transmission would have occurred successfully to allow HIV or a close progenitor to enter the human population during the Congo poliovirus vaccine trials.

1) Contamination of the vaccine with SIV or similar viruses (hereafter referred to as SIV). SIV does not multiply in or infect epithelial or fibroblastic cells from monkey kidneys and neither does HIV. However, poliovirus for the vaccine was grown in primary cultures that could have contained a low number of lymphocytes and macrophages, known to harbor SIV in vivo and to support the replication of SIV in culture. Thus, the possibility of the presence of a small amount of SIV particles in these culture supernatants cannot be discounted. Such cell culture-derived contamination could have taken place at any time during propagation of the attenuated virus stocks, including the preparation of "seed" virus from which the lots actually used in the field were derived by subsequent passages in monkey kidney cultures. We consider it, however, quite unlikely that the contaminating SIV was picked up during the first passages of poliovirus and carried along with poliovirus in all subsequent passages, since the condition of infection of tissue culture cells were not favorable to SIV propagation. It is more likely that contamination (if any) would have occurred during the preparation of the vaccine lots actually used in the field trials. The probability that SIV may be present in kidneys and kidney cultures from monkeys infected with SIV has actually been addressed in a recent publication. Ohta et al. (AIDS, 3:183-185, 1989) found no evidence for the presence of infectious SIV particles in kidney cell cultures from SIV infected monkeys, although the virus could be recovered from peripheral blood lymphocytes, spleen and other organs. It is possible that more sensitive methods could have detected SIVs in these cultures, but clearly this evidence indicates that the extent of contamination, if any, should have been small.

A crucial point here is the origin of the monkey kidneys used for growing the vaccines. Were they derived from Asian or African monkeys? Wild-caught macaque monkeys from Asia are believed to be SIV-free, and were generally used at the time. If the vaccines had been prepared from tissue of these animals, the probability that they had been infected could be considered very low. Around the time of the vaccine trials, however, the Indian government put an embargo on monkey exports and thus monkeys of African origin may have been used. Were the kidneys used for preparing the vaccine derived from macaques or African green monkeys (AGM)? This is important because AGMs are known to harbor SIV naturally in the wild, with a prevalence of about 30%. Unfortunately, the origin of the kidneys used in the preparation of the 1957 vaccine is unlikely to ever be determined with any certainty.

The last point to be considered in this hypothetical first step of transmission is whether SIVs would have survived the treatments used in the preparation of the vaccines. While poliovirus is very stable, HIV and SIV, like other retroviruses, are quite labile in the extracellular environment. The polio vaccine was subjected to at least two cycles of freezing and thawing (Wistar Institute protocols), a procedure known to cause significant loss of SIV and HIV infectivity, while reducing poliovirus titers only marginally. In addition, poliovirus vaccines were kept for an indeterminate length of time at -20^oC, a condition also known to impair retrovirus infectivity. Furthermore, the poliovirus stock used in Congo was diluted with salt solution approximately 300 fold to obtain the final vaccine preparation.

In summary, the possible presence of SIV or related virus particles in the vaccine preparation cannot be discounted. However, if present, the concentration of SIV particles is likely to have been extremely low.

2) Administration. In the Congo trial ~1 ml of vaccine was administered to each individual by squirting into the oral cavity with the use of an automatic syringe. About 300,000 individuals were vaccinated. The oral route is not an efficient was of infection by SIV or HIV, which are generally transmitted through blood or sexual contact. It is possible, however, that some of the vaccinated individuals had sores, wounds or blisters in their oral cavity, thus increasing the chances of infection. In addition, there is evidence that the virus can be transmitted on rare occasions from mother to newborn via breast-milk, suggesting an oral or gastrointestinal route of transmission in infants. In any case, the chances of SIV or HIV being successfully transmitted through the method of administration used for the polio vaccine are low.

The vaccine trials conducted in Africa were carried out mostly by squirting the vaccine aliquots through an automatic syringe into the mouths of recipients. In contrast, the Sabin attenuated poliovirus vaccine, which was also produced in monkey kidney cells, was distributed to millions of recipients in Eastern Europe, in sugar cubes. Considerable emphasis has been given by Curtis to the different methods of distribution as a possible explanation for the fact that the AIDS epidemic did not start in Eastern Europe. We find this argument to carry little weight. Sucking a sugar cube would give a virus particle approximately the same likelihood to reach a susceptible cell or wound in the oral cavity as mouth squirting of the same amount of fluid. Aerosolization could conceivably increase the chances of the virus reaching the respiratory tract and alveolar macrophages, but aerosolization deep into the lung is not likely to have been common.

3) Multiplication in human subjects of the putatively transmitted monkey virus. If SIV or an SIV like virus did gain access to a susceptible cell in a vaccine recipient, it is possible that it might have multiplied in the human host. There are two known types of human immunodeficiency virus, HIV-1 and HIV-2. HIV-1 is the virus prevalent in central Africa, U.S., and Europe; it is also responsible for most cases of AIDS worldwide. In contrast, HIV-2 is less prevalent and is largely confined to west Africa. HIV-2 is closely related to SIV of sooty mangabey monkeys and is generally believed to have arisen by cross-species infection from monkeys to humans in west Africa. Indeed, a recent report (Khabbaz et al., Lancet 340:271-273, 1992) has shown that SIV has infected at least one laboratory worker, consistent with laboratory observations that SIV can grow in human cells. However, a retrovirus closely related to HIV-1 has never been identified in monkeys.

4) Bridging the evolutionary distance between SIVs and HIV-1. Although HIV-1 and the known SIVs from green monkeys, sooty mangabeys and mandrills are evolutionarily related, their nucleotide sequence homology is only about 40%. It is the view of many experts in the field that the evolution of SIV into HIV-1 was a long process, measurable more in decades or centuries than in a few years. SIV from sooty mangabey monkeys and HIV-2 are more closely related, but the predominant virus in the AIDS epidemic is HIV-1. The earliest evidence of HIV-1 infection date to 1959 (a seropositive individual in Zaire, and a seaman who died in 1959 with typical AIDS symptoms and later found to have been infected with HIV-1). It is extremely unlikely that the evolutionary gap between HIV-1 and any known SIVs was bridged in less than two years. Moreover, we have investigated further details of the earliest documented case of HIV-1 infection, the merchant marine mentioned above, who was symptomatic in 1958 and died of AIDS in 1959 in Manchester, England (G. Williams et al. Lancet, 1960; ii:951-955; Corbitt et al. Lancet, 1990, 336:51). It appears that this man travelled abroad to Gibraltar and Northern Africa beginning in 1955. However, he had returned to England by the first half of 1957 before the Congo trial was begun. Therefore, it can be stated with almost complete certainty that the large poliovaccine trial begun late in 1957 in Congo was not the origin of AIDS.

5) Testing of the vaccine samples used in the Congo trials. Various articles in the press have declared that the Curtis' hypothesis could be easily verified by testing aliquots of the vaccine(s) used in the Congo trials. As stated above, testing of a master seed or related vaccines will not be sufficient, as the contamination with SIV (if any) would be likely to have occurred during the preparation of the actual vaccine lots used. Thus, the vaccine samples will have to be conclusively identified as the ones used in the Congo trials or at best lots prepared at the same time in the same laboratory. If these can be found, they can be tested by attempting recovery of infectious virus or by reverse transcriptase-polymerase chain reaction (RT-PCR). Virus infectivity is almost certain to have been lost, since these samples have been stored for 35 years, possibly subjected to cycles of freezing and thawing, etc. Thus, a negative result will be inclusive. PCR amplification would have to be performed after reverse transcription, using primers bracketing conserved regions in the SIV/HIV genomes. Again, a negative result may not be conclusive and a positive result will not identify the virus type. Thus cloning and sequencing of the PCR products would be required. Finally, the significance of the presence of SIV sequences only distantly related to HIV-1 may be difficult to evaluate.

It is also to be considered that attenuated poliovirus vaccine was universally prepared in monkey kidney cells at that time and even at much later times (including today). Thus a thorough analysis would require that a number of vaccine lots used during the late 50's and early 60's should also be tested. Millions of individuals were vaccinated in Russia and Eastern Europe during those times with vaccines prepared similarly. There is no evidence that the AIDS epidemic began in these parts of the world. What if a significant percentage of the polio vaccines used there were also found to be positive for SIV-like viruses?

In conclusion, testing of the vaccines will be laborious, expensive and may be inconclusive.

Summary and recommendation

It is apparent that we consider the probability of the AIDS epidemic having been started by the inadvertent inoculation of an unknown HIV precursor into African children during the 1957 poliovirus vaccine trials to be extremely low. Almost every step in this hypothetical mode of transmission is problematic. The contamination of the poliovirus vaccine lots with SIV/HIV particles, if any, is likely to have been extremely small. Transmission by the oral route is extremely rare for HIV or SIV. Finally, the evolutionary distance between known monkey immunodeficiency viruses and HIV-1, the prevalent virus in the AIDS epidemic, probably took decades or centuries to be bridged and not a few years. The most telling evidence is the case of the Manchester sailor who appears to have been infected with HIV-1 even before the poliovirus trials were begun in Congo.

Such considerations notwithstanding, we feel that some testing of available samples may be desirable so that no stone is left unturned. We are pessimistic that such testing, even if performed by the best labs using the most appropriate techniques, will produce conclusive results. Nevertheless, a limited number of samples of vaccine stocks available at the Wistar Institute could be tested for the presence of HIV/SIV viruses or related nucleic acid sequences. Of the samples presently existing at the Wistar Institute, only one has been identified as being possibly directly relevant to the Congo trials. Several other samples were prepared around those times, but may never have been used in the 1957 vaccine trials. In sum, the total number of samples (also considering the amount of material available) worth testing is quite limited.

Testing should be done by, or under the direct supervision of, the World Health Organization (WHO) or the U.S. Centers for Disease Control (CDC). Coded samples should be tested in at least two experienced laboratories. Testing for the presence of infectious SIV or HIV particles should be done using the following target cells:
- Human peripheral blood lymphocytes cultures
- African green monkeys PBL cultures from SIV negative Caribbean Monkeys
- the MoLT-4 clone 8 cell line
- the HUT78 cell line

These cell lines should be capable of detecting most or all of the known HIV/SIV species. Virus production should be monitored by testing for reverse transcriptase activity in the culture medium, and by an antigen capture assay. Any virus produced should be fully characterized.

Testing for the presence of HIV/SIV related sequences by PCR should be done on nucleic acid extracted from vaccine stocks, possibly after a step of virus concentration. Reverse transcriptase - PCR should be used. Primers for the amplification of HIV/SIV related DNA should be the following.
a) Two HIV-1 specific primers (representing conserved regions of the viral genome)
b) Two HIV-2 specific primers
c) SIV^SM-SIV^MAC specific primers
d) SIV^AGM-specific primers
e) Two primers which could detect sequences conserved in all members of the HIV family. The pol gene is a good candidate for this.
Any amplified DNA should be sequenced. Stringent measures should be adopted to avoid laboratory contaminations

In closing, we feel compelled to mention that the current controversy highlights the problems and difficulties associated with using monkey tissue for production of vaccines administered to humans. To this day, live-attenuated poliovirus vaccine is produced in the United States and in most other countries using primary African green monkey kidney cells. Although green monkeys can now be certified free of SIV for use in vaccine production, specific tests could not have been performed prior to 1985 when SIV was first isolated. There may well be other monkey viruses that have not yet been discovered that could possibly contaminate vaccine lots. This provides a powerful argument for the use of well-characterized cell lines for vaccine production. A continuous cell line can be intensely scrutinized prior to certification for use in vaccine production. It could be argued that the live polio vaccine produced in primary monkey kidney culture is extremely safe and effective and that tampering with the mode of production could result in diminished safety or efficacy. However, at least one major European manufacturer has recently gone to using a cell line and is apparently producing a safe and effective live vaccine with this method. A serious effort is needed in the U.S. and other countries to effect a switch to well-characterized cell lines for vaccine production.

Claudio Basilico
Professor and Chairman
Department of Microbiology
NYU School of Medicine
New York, NY 10016

Clayton Buck
Professor
The Wistar Institute
Philadelphia, PA 19104-4268

Ronald Desrosiers
Professor of Microbiology and Molecular Genetics
Harvard Medical School
Chairman, Division of Microbiology
New England Regional Primate Research Center
South Borough, MA 01722-9102

David Ho
Director, Aaron Diamond AIDS Research Center
New York, NY 10016

Frank Lilly
Professor, Department of Molecular Genetics
Albert Einstein College of Medicine
Bronx, NY 10461

Eckard Wimmer
Professor and Chairman
Department of Microbiology
State University of New York at Stony Brook
Stony Brook, NY 11794-5222