Preliminary Notes Concerning Shortcomings of a
Correspondence by Y. Ohta, et al., Entitled “No Evidence for the Contamination of Live Oral Poliomyelitis Vaccines with Simian Immunodeficiency Virus,” Published in AIDS, 3: 183-4, 1989
Introduction: This piece is to be read in conjunction with L
Pascal, What Happens When Science Goes Bad, Science and
Technology Analysis Working Paper #9, Department of Science and
Technology Studies, University of Wollongong, Wollongong, NSW 2522,
Australia, Dec 1991. It should also be read in conjunction with the
report of the Wistar Committee (C Basilico, C Buck, R Desrosiers, D
Ho, F Lilly, E Wimmer, “Report from the AIDS/Poliovirus Advisory
Committee,” Wistar Institute of Anatomy and Biology, 3601 Spruce St.,
Philadelphia, PA 19104, 18 Sep 1992). The Ohta paper is discussed at
some length in the appendix to What Happens When Science Goes
Bad, but its errors greatly exceeded the space available to point
them out. In addition, both the Wistar Committee and Koprowski have
continued to make important use of the Ohta piece in their arguments;
hence this more detailed treatment.
For those who have not read What Happens When Science Goes Bad,I include here a synopsis, together with some important developments that have occurred since its publication.
My piece presented detailed arguments that HIV-1 was transferred from its simian hosts into human beings through contaminated oral polio vaccine made from monkey kidney cultures infected with the simian precursor virus of HIV-1, just as the simian virus SV-40 was transferred into millions of human beings through many contaminated batches of polio vaccine and just as several other monkey viruses are known to have contaminated that vaccine and other vaccines. The difference is that the millions infected with SV-40 nearly all caught it from the vaccine itself, since many batches were contaminated and since the virus does not ordinarily spread from one person to the next, while the millions infected with HIV-1 nearly all caught it from other infected individuals, since it appears that only one batch was contaminated with HIV-1’s precursor, and since only a small fraction of the 300,000 receiving this contaminated batch would have become infected via the oral route. This batch of vaccine was made by Hilary Koprowski at the Wistar Institute and was used in Central Africa in 1957-58: this African campaign was in fact the world’s first mass oral polio vaccination campaign.
Wistar has failed to respond to requests for samples of Koprowski’s vaccine for testing (either that batch, which they say they are not sure they can find, or other early batches made by him, which they have said they see no reason to supply). Wistar appointed a 6-person committee to study the question, which, after more than half a year, produced an undocumented, 7-page, 2600-word report concluding the Wistar vaccine was extremely unlikely to have been the cause of AIDS, but nevertheless recommending discontinuance of vaccine manufacture in monkey kidneys. (For comparison, the preliminary piece you are now reading, not counting this introduction, is 4200 words long. It was done by one person in a few days, with all of its controversial points documented, and it discusses only a tiny aspect of the problem.) The Wistar Committee called for testing of vaccine samples and suggested two of the organizations most closely linked to the coverup to do the testing (together with saying in interviews that samples should be denied to Robert Bohannon, who became suspicious after finding the first reported case of another monkey immunodeficiency virus in an AIDS patient–see RC Bohannon, et al., Journal of Virology, 65: 5663-72, 1991–and whose requests to Koprowski and the FDA for samples for testing were not even acknowledged by Koprowski and were met by one stall after another from the FDA). Wistar has said it will accept the committee’s recommendation. If it has supplied any samples for testing, I have not heard of it; but why should anyone believe results obtained through such a brazenly cooked procedure anyway? The case for AIDS’ origin has been blocked from publication in mainstream scientific journals, at least one of which has attempted to justify its decision to withhold this information from the world. (See the editorial “A Startling 19,000-Word Thesis on the Origin of AIDS: Should the JME Have Published It?” by Raanan Gillon in Journal of Medical Ethics, 18: 3-4, 1992.) A piece by investigative reporter Tom Curtis in 19 March 1992 Rolling Stone, and many follow-up stories by him in the Houston Post and elsewhere, resulted in a lawsuit for defamation by Koprowski against Curtis and Rolling Stone, which is now pending. Curtis developed his ideas in collaboration with Blaine Elswood and Raphael Stricker, all of whom were unaware that they were duplicating my prior work in this area.
(Note: The body of this piece has not been written yet. What follows is a summary of the major points I will make, with, however, some of the argumentation and documentation included here. In the finished version most of these latter will be transferred to the body of the piece. I have written this preliminary account because it was something I could do quickly, and I do not know when I will have time to do a more complete job.)
In summary: The piece by Ohta et al. cannot be used to support the contention that Koprowski’s monkey kidneys did not transfer AIDS’ ancestor into human beings through polio vaccine because:
1) They tested the wrong virus. Different simian immunodeficiency viruses (SIVs) vary by about 40 percent in their nucleotide sequence, and nobody at all thinks HIV-1 came from the common SIV of African green monkeys. Ohta found no SIV in the livers of the monkeys he tested. Yet YZ Cao, et al., AIDS, 6: 65-70, 1992, found high levels of HIV-1 in the livers of several of 14 AIDS patients. Perhaps the difference in these results is accounted for by the fact that we are dealing with two different viruses, and perhaps the list of tissues SIVagm will grow in does not hold for HIV-1, or for its simian precursor. Or perhaps the difference is due to the fact that we are dealing with two different species (see 2, below). Or perhaps the difference is due to the fact that Ohta et al. tested the livers of only two monkeys. Cao et al. found no detectable HIV-1 in the livers of 5 of the 14 AIDS patients, thus showing the dramatic differences that can occur with different samples even of the same immunodeficiency virus, let alone one whose genetic sequence differs by 40 percent. Perhaps if Ohta et al. had tested the livers of more than two monkeys, they would have found some containing high virus levels also. Perhaps if Ohta et al. had tested the kidneys from more than two monkeys, they would have found some containing high virus levels also. The Wistar Committee can hardly claim to have been unaware of the dramatic differences found among different samples even of the same immunodeficiency virus, because one of its six members, David Ho, coauthored the Cao paper. Indeed, this and similar points are too well known for any members of the committee to have been unaware. When different isolates of HIV-1 are taken even from the same individual, they differ markedly in what cell cultures they will grow in: see C Cheng-Mayer et al., Science, 240: 80-82, 1988. In only 8 samples taken from only 3 individuals, Cheng-Mayer found several different pairs of samples where one of the pair would grow in A but not B, while the other would grow in B but not A. For example, isolate SF665 would grow in primary macrophages but not in Jurkat, CEM, or U937 cells. Isolate SF118 would grow in Jurkat, CEM, and U937 cells but would not grow in primary macrophages. Yet we are asked to believe results obtained from only three samples tested in only two types of cells, one of these two being cells of the wrong species (human lymphocyte cultures). To believe results obtained from such a little bit of testing of such a changeable virus would be nonsense even if the virus they had been testing had been the right one.
2) According to Koprowski, they not only tested the wrong virus but tested it in the wrong species. Koprowski says he did not use African green monkeys. However, nobody knows whether Koprowski is right, inasmuch as he has made several contradictory claims as to what monkeys he used.
3) The kidney cultures Ohta et al. used were not superinfected with poliovirus. That would change the environment substantially. More specifically, they were not superinfected with the unusual strain of poliovirus Koprowski’s kidney cells were infected with.
4a) Their kidney cultures were not superinfected with the simian retrovirus MPMV or related D type retroviruses or foamy retrovirus or STLV-1 retrovirus, or various herpesviruses related to EBV and CMV, or SV-40, or other known and unknown viruses, any one of which might have enabled HIV-1’s ancestor to infect cells it could not normally infect, just as they can enable HIV-1 to infect mouse cells, and HeLa cells, and other non-CD4 cells, which HIV-1 cannot normally infect (see, e.g., B Chesebro, et al., Journal of Virology, 65: 5782-89, 1991).
4b) They did not test these viruses in combination. This test, of all viruses, known and unknown, common and rare, in all combinations, cannot possibly be run. And without it the claim that monkey kidneys can not support any particular virus is worthless. Perhaps when SV-40 infection occurs together with simian cytomegalovirus infection and STLV-1 infection, then ordinary kidney cells with no lymphocytes at all become a wonderful medium for growing SIVs. Perhaps the reason only one batch was apparently contaminated is that this particular combination of four simultaneous contaminations occurred in only that one batch of all the thousands that have ever been made. Or perhaps the contaminating virus discovered in that batch by Sabin and never identified is a rare virus that infected only this one batch and it is what enabled AIDS’ ancestor to grow there. There are innumerable such possibilities and no way to test for more than a tiny fraction of them. All indications are that the batch was contaminated with the ancestor of HIV-1. Objections to this view are based on theoretical reasons why the SIV would not have grown in kidney cells. Yet here is a large, unsealable hole in those theoretical reasons sufficient to invalidate them even if every other test had been done perfectly and had produced unequivocal results.
5) They used very insensitive tests when far better ones were available. Reverse transcriptase assays and immunofluorescence tests are not nearly sensitive enough to detect low levels of infection. Far better tests, such as the PCR, exist but were not used, the simplest being to inject kidney cultures from infected monkeys into noninfected monkeys to see if they develop the infection.
6a) They did not allow for the fact that HIV-1 concentrations may be thousands of times greater than normal in a few rare individuals. Wistar Committee member David Ho, in New England Journal of Medicine (321: 1621-5, 1989) found 2 of 54 individuals who had more than 1600 times as much virus as the average for asymptomatic individuals and who had 10,000 times more virus than the 9 individuals with the lowest totals. If monkeys infected with SIV show comparable ranges, the results might have been very different if Ohta had tested enough of them to find one with the peak value. By what strange stretch of the imagination can David Ho have assented to the committee’s conclusion that Ohta’s letter had shown that any possible contamination would have had to be very low level: multiply a low figure by 10,000, and it may not be so low any more.
6b) This enormous variation in virus concentration is by no means produced just by differences among the individuals infected. Different strains of the same immunodeficiency virus would show great variation even if they were tested in identical individuals. This would be a difficult experiment to run, and to my knowledge it has not been run. Nevertheless, very strong evidence for this contention can be gotten through indirect means, such as is presented in B Spire, et al., Gene, 81: 275-84, 1989. These researchers discovered an African isolate of HIV-1 that in 7-day tests caused cytopathic effects (i.e., visible damage) in cultures of MT-4 cells even though it was 10,000 times more dilute than the maximum dilution at which standard HIV-1 strains caused damage. MT-4 cells are the same ones Ohta, et al. say they used to test for SIVagm in monkey kidneys (but see below for a caveat). Still more evidence for the variability of different strains of the same immunodeficiency virus is given by ES Daar, et al., Proceedings of the National Academy of Sciences, 87: 6574-78, 1990. These researchers found great variation in the amount of soluble CD4 necessary to neutralize different HIV-1 strains: they isolated virus from only 10 patients infected with HIV-1, yet found a 14-fold variation among these 10 strains, the most resistant of which was able to withstand 2700 times as much CD4 as typical laboratory strains. The Wistar Committee cannot claim to have been unaware of these great differences among strains, because one of the coauthors of this piece was Wistar Committee member David Ho.
6c) This enormous variation in virus concentration is by no means produced just by differences among the individuals infected and differences among viral strains. There is also a very great variation in the amount of virus present over time. Among infected human beings the quantity of virus present is small for most of the period of infection but is very large for the first few weeks after infection and very large after symptoms develop. ES Daar, et al., New England Journal of Medicine, 324: 961-5, 1991 found that in 4 very-recently-infected patients the quantity of HIV-1 in plasma decreased from the peak values attained during the initial infection by a factor of 1000-fold over the course of just 9 days in one, by 1000-fold over the course of 5-1/2 weeks in another, by 1000-fold over the course of 6 weeks in another, and by 10,000-fold over the course of 6 weeks in the remaining patient. If infected monkeys behave like infected people, most will have quite low amounts of virus most of the time, but they may have 10,000 times greater amounts during brief periods, especially when initially infected. Perhaps if Daar et al. had tested more than 4 patients, they would have found even greater variation. Again, one of the coauthors of this piece showing such enormous time variation is Wistar Committee member David Ho.
6d) So what level of virus should we expect if we took one of the monkeys most prone to producing high viral levels, infected him with one of the highest-producing viral strains, and took his kidneys just as the peak level was reached? Such levels might far exceed 10,000 times typical values. Such a coincidence might not happen very often, but when we have vaccine being made from thousands of monkeys over the years and around the world, it is inevitable that such an unfortunate confluence of circumstances will eventually be encountered.
7) Ohta et al. attempted to prove a negative conclusion–that SIV never exists in monkey kidney cells–by running only two experiments on two monkeys each, when thousands of monkeys have been used in manufacturing vaccine, and when, from sequence evidence, it can be stated with near certainty that only one monkey among these thousands was responsible for today’s epidemic of HIV-1. There are many possible explanations for why only one monkey was responsible: perhaps he was infected with a virus from a different simian species, as happened with rhesus and cynomolgus monkeys accidentally infected with SIV from sooty mangabeys, and because his species’ immune system had never evolved to contend with this virus, it was unable to keep it in check; or perhaps he was born with a naturally weak immune system that could not keep the virus in check; or perhaps his immune system was weak due to the fact that he was very old, or very young, or very hungry, or very stressed by the conditions of his capture and shipment; or perhaps he was coinfected with other viruses which weakened his immune system and/or allowed HIV to infect non-CD4 cells; or perhaps he had a coexisting kidney infection leading to increased numbers of lymphocytes in his kidneys; or perhaps he had a strain of virus that grew to higher titers everywhere, or to higher titers in kidney cells, or was better at infecting non-CD4 cells, or better at growing in tissue culture, etc., etc., or various of these factors in combination–see 6d, above. If only one monkey among these thousands had detectable levels of SIV in his kidneys, then if Ohta et al. had tested an equal number of monkeys (i.e., many thousands), they would have had only about a 50 percent chance of finding such a case. Yet they are attempting to argue that such a thing never happened by testing a grand total of only four monkeys. And even this 50 percent chance depends on their running good experiments on these thousands of monkeys.
8) Both these experiments were severely flawed. In their first experiment, they did not allow for the fact that the virus might be in a latent state, was unadapted for growth in human cell cultures, and that different isolates even from the same individual monkey, let alone different monkeys, let alone different species of SIV from different species of monkeys (which applies in the case of the Wistar Committee’s citation of this piece), differ dramatically in their ability to infect particular culture media. In their second experiment they did not allow for the fact that they were testing in media which they themselves say contained only “a few” lymphocytes (presumably far less than 1 percent), and therefore that SIV levels might be greatly reduced in comparison with media consisting of 100 percent lymphocytes, nor did they lengthen the period of the experiment to accommodate for the lack of lymphocytes. This experiment also used SIV adapted for growth in human cells, and an additional period for readaptation to monkey cells might have been needed.
9a) Their tests were equivocal at best and may actually have indicated SIV was present and growing. There is a vast no-man’s land between a finding of no SIV and a finding of SIV, and in this no-man’s land the results are indeterminate: their experiments were indeterminate, with whatever inconclusive evidence that was there pointing toward the presence of SIV, not its absence.
9b) As soon as the experiments began to look as if they might be indicating the presence of SIV, they were for no possible scientific reason halted: look at the growth of SIVagm in infected TALL-1 cells in Fig. 5b of Y Ohta, M Hayami, S Honjo, H Tsujimoto, K Ishikawa, et al. (all coauthors of the Ohta letter), International Journal of Cancer, 41: 115-122, 1988: the SIV does not show detectable growth for the first 20 days, and at 24 days it looks very much like the second graph in Ohta’s AIDS piece looked at 28 days, at which point that experiment was halted. But at 36 days, when the TALL-1 graph ends, it is ahead of some other clearly positive cultures and is continuing upwards with its greatest slope so far.
10) They say: “From these results, poliomyelitis vaccines may be considered not to be contaminated with SIVagm, even though they are prepared in primary kidney-cell cultures from SIVagm-infected AGM.” People who run only two experiments on only two monkeys each and then proclaim the medium is safe enough to risk the lives of millions of people are crazy, regardless of how good a job they have done in these two experiments or how conclusive the results might have been. This would be true even if there were not great amounts of evidence pointing to this medium as already having given us AIDS. It is not a good idea to stake millions of lives on the quality of the work of obviously crazy people, and those who do so are also crazy.
11) Two of the eight authors are shown as working for the Japan Poliomyelitis Research Institute, while another is at the Tsukuba Primate Center for the Medical Sciences. The remainder are at various Japanese medical facilities. All of them had a strong vested interest in coming to the conclusions that they did. All would have been risking their careers had their results been anything else.
I have just listed 11 fatal shortcomings of this paper. They are fatal in the sense that any one of them, even if everything else had been done correctly, nay outstandingly, would have been sufficient to disqualify the paper from the role the Wistar Committee has assigned to it. Of these 11 fatal errors, numbers 1, 2, and 11 are errors of the Wistar Committee in making inappropriate use of a paper that was inadequate for their purposes. The other 8 are errors inherent in the paper itself. Unless and until someone comes up with a paper containing more than 8 separate fatal errors (or until someone shows that these are not all errors, or are not all fatal, or are not all separate), I am claiming this for a world’s record: the worst paper ever published in the field of science. It is unfortunate that this worst paper ever, or strong contender for worst, should have been published on what is the most important subject ever investigated by science, or strong contender for most important, and that it should have been given the key role it has been given: this paper is the
only documentation the Wistar Committee presents for any controversial point, some of which points, such as their claims about the British sailor’s travels, contradict published information (e.g., P Brown, New Scientist, 127(1725): 34, 14 July 1990). The committee has nothing. They have made a mockery of science and of common sense. Yet they propose we stake millions of lives on their work.
Beyond this, Ohta’s piece contains a confusing error or set of errors I do not know how to interpret. In the text, they say the first experiment was done by attempting to infect MT-4 cells using kidney material from infected monkeys. Yet the graph (Fig. 1) shows reverse transcriptase levels in infected and uninfected MOLT-4 cells. Despite their similar names, MT-4 cells and MOLT-4 cells are not the same. Which did they use?
Text should ordinarily take precedence over figure captions. This is especially so since the text explicitly states that MT-4 cells were being used, while Figure 1, which shows levels for infected and uninfected MOLT-4 cells, does not explicitly say that MOLT-4 cells were being used. Taken by itself, Figure 1 strongly implies that MOLT-4 cells were used, but taken together with Figure 2, there is no such implication. Infected and uninfected MOLT-4 lines are also drawn in Figure 2, when it is very clear that here at least they were attempting to grow the virus in primary monkey kidney cells, rather than either MOLT-4 or MT-4. It appears they mean the MOLT-4 lines to be no more than a generic representation of reverse transcriptase activity levels in cultures before and after infection. This interpretation removes the apparent conflict between text and figure, but it still suffers from the problem of why the levels in both experiments are lower than background. But this problem cannot be given as a reason for favoring figure over text, because the same problem exists whether they used MT-4 or MOLT-4 cells. (Their choice of MOLT-4 and the particular SIVagm isolate they used to infect it can also be challenged as inappropriate for illustrating before and after levels, but I will not do that here.)
If MT-4 cells were used for this test, as they explicitly say in the text was the case, then we need to know why a cell type that these same authors have shown in earlier experiments does not reliably detect SIVagm was used. Quoting from a paper coauthored by two of the authors of the letter in question (Ohta and Hayami), as well as by Ronald Desrosiers of the Wistar Committee (MD Daniel, et al., Journal of Virology, 62: 4123-28, 1988):
“SIV from these African green monkeys (SIVagm) was isolated and continuously produced with the MOLT-4 clone 8 (M4Cl8) cell line but not with a variety of other cells including HUT-78, H9, CEM, MT-4, U937, and uncloned MOLT-4 cells. Once isolated, these SIVagm isolates were found to replicate efficiently in M4Cl8, SupT1, MT-4, U937, and Jurkat-T cells but much less efficiently if at all in HUT-78, H9, CEM, and MOLT-4 cells. [I.e., once, from among the considerable variation existing within any single monkey, a viral strain had been isolated that could grow in MOLT-4 clone 8, and once it had had a little time to adapt to growth in this particular human lymphocyte culture, it then became able to grow efficiently in other human lymphocyte cultures that previously would not suffice, yet even so would not grow in normal MOLT-4, thus showing how much even the same type of culture cells can differ from one clone to the next, let alone kidneys from different monkeys, let alone from different monkey species.] The range of CD4+ cells fully permissive for replication of these SIVagm isolates thus differs markedly from that of previous SIV isolates from macaques (SIVmac).” [Emphasis mine]
(This quote is from the abstract. Later in the same paper, p. 4124, they describe the difference between SIVagm and SIVmac in even stronger terms: “The CD4+ cell lines that support replication of these SIVagm isolates differ dramatically from those that support replication of SIVmac.” Since at least one member of the Wistar Committee, co-author Desrosiers, was aware of the dramatic differences between SIVs from different monkeys, since indeed he has authored several papers on the differences among SIVs, and indeed is one of the foremost authorities on the issue, how in the world could the committee use the results Ohta obtained for a type of SIV that no one at all believes was the parent of HIV-1, even if the paper were exemplary in every respect?)
Well, perhaps they used MOLT-4 cells rather than MT-4, as one would gather from looking at Figure 1 and not looking at Figure 2 and not looking at the text. Then we must ask if this was MOLT-4 clone 8, which detected SIV in the few monkeys they tried it on, or uncloned MOLT-4, which did not? If they were attempting to do a test which would detect SIVagm if it was present, then they used clone 8. If, on the other hand, they were attempting to run tests that would fail to detect SIVagm so as to cover up polio vaccine’s role in the origin of AIDS, then they did not use clone 8. If they were doing their best to reach the truth, then the list of shortcomings I have presented shows them to be quite unbelievably incompetent. But if instead they were attempting to cover up the truth, then many of the “gross errors” become clever subterfuges, and much of the incompetence disappears.
Consider: They write a grand total of 5 paragraphs, during which they make 8 fatal errors, and it is not even clear what cells they tested in, while the ones they say they tested in had already been shown by the authors themselves (and a member of the Wistar Committee) to have been totally inappropriate. Yet the committee picked by the Wistar Institute is so intent on finding something, anything, that can absolve Wistar’s vaccine from having started AIDS that they overlook every one of these 8 fatal errors. But even without these fatal errors, the paper is still inappropriate for their use. They must make 3 more fatal errors of their own in order to use it. This they do. And these 5 paragraphs, which have now been responsible for 11 separate fatal errors, are the only reference the Wistar Committee cites for any controversial point. I do not know whether these errors were accidental or deliberate. But when one interpretation leads to unbelievable incompetence and the other causes otherwise unexplainable errors suddenly to make sense, it becomes highly suspicious, to say the least. Either way, whether we are in the hands of world-record fools or Machiavellian mass murderers, we had better find some way to escape their clutches.
There is an important lesson to be learned here: If you allow yourself enough errors, you can come to any conclusion you want. This is precisely what was done by Ohta, his coauthors, and the Wistar Committee. They simply ignored everything that stood in their way. They used the wrong virus and may well have used the wrong monkeys. They apparently used culture cells they themselves had shown would not work, then used kidney cells uninfected with poliovirus, likewise uninfected with other simian viruses that commonly contaminate kidney cultures, some of which are known to increase the range of cells in which SIVs will grow. They made no allowance for the fact that they themselves had shown SIVs are both extremely variable and extremely sensitive to different culture methods, made no allowance for the fact that they themselves had shown viral levels may vary by factors of at least 10,000-fold, even in a single individual at different times. They used insensitive detection methods, ran only two experiments on two monkeys each, the design of which experiments was badly flawed, proclaimed the results were negative when they were indeterminate, proclaimed the experiments over when it began to look as if they might detect SIV after all, proclaimed they had shown modern polio vaccines were not contaminated with SIVagm, proclaimed this implied Koprowski’s vaccine was not contaminated with the precursor of HIV-1. And all this was done by people with a great deal to lose, representatives of parties with a great deal to hide, parties whose incompetence and irresponsibility back in 1957-58 unleashed on humankind the worse disaster it has ever faced, parties who are now being accused of exactly that, and who in addition are being accused of covering up the evidence for eight long years.
And through it all the monkeys, who are subject to other immunodeficiency viruses, and many dangerous viruses of other kinds, have continued to be used in making vaccines. A few months ago, on 14 December 1992, William Haseltine of Harvard testified before Congress that more than a billion human beings would be infected with HIV-1 by the early decades of the next century. And how many in the century after? And how many in the century after that? And how many other diseases will we transfer in the meantime in the very same way?