AIDS and the Polio Vaccine
Letter to Science by Hilary Koprowski
Adapted from Science, vol. 257, 21 August 1992, pp. 1024, 1026-1027; corrections, 11 September 1992, p. 1463. In this version, the errors in footnoting have been rectified so that the footnotes read as they were intended.
As a scientist, I did not intend to debate Tom Curtis when he presented his hypothesis about the origin of AIDS in Rolling Stone (1). The publication of his letter in Science (29 May, p. 1260), however, transferred the debate from the lay press to a highly respected scientific journal. I would now like to state my views, based on facts, in order to counter and thereby repudiate Curtis' hypothesis about the origin of AIDS.
Polio epidemics were raging throughout the world in the late 1940s. At that time, the late Tom W. Norton and I developed a concept to try to attenuate the poliovirus through its passage in unnatural nonprimate hosts and through cloning of the virus for selection of variants avirulent for monkeys injected intracerebrally. It was a difficult task but we finally succeeded (2), between 1949 and 1952, in attenuating the three types of poliovirus and using them for vaccination in preliminary trials in the United States.
Curtis has theorized (3) that the "African epidemic was spawned by a contaminated polio vaccine administered from 1957 to 1960 to at least 325,000 people in Rwanda, Burundi and the former Belgian Congo." He has stated that the area of vaccination of children in Ruzizi Valley in 1958 corresponds to "roughly to another map . . . the one identifying the regions of highest HIV [human immunodeficiency virus] infection in equatorial Africa" (1). This is completely wrong. Ruzizi Valley, where 215,504 subjects were vaccinated in 1958, is located in the northwestern part of the Republic of Burundi, not in the Kivu district of Zaire, an area where Curtis placed "the lion's share of their [Koprowski and his associates] samples" (1). Curtis refers to the "high prevalence of antibodies" to the AIDS virus without symptoms of disease in the Kivu district. The finding of a high rate of HIV among healthy populations (4) could be ascribed to the well-recognized fact that ELISA (enzyme-linked immunosorbent assay) tests of the first generation developed at that time show a high rate of nonspecific positive reactions. This was taken into consideration in the second report of the same investigators (5), who say that "reactivity in both ELISA and Western Blot analysis may be non-specific in healthy Africans" (6). Thus, the high prevalence (12 to 24%) of HIV in the rural Kivu population had to be scaled considerably downward, particularly in the light of recent reports that out of 675 blood donors from Kivu only 25 (3.7%) were seropositive for AIDS antibodies.
Since 1985 many epidemiologic studies have been carried out in the rural populations of northeast Zaire, Burundi (Ruzizi Valley), and Rwanda. The results of these studies confirm the previously published low rate of HIV infection in these areas [0.7% for rural Burundi (7), 1.3% for rural Rwanda (8), and 3.7% (6) for rural Zaire]. It is thus misleading for Curtis to say that Ruzizi Valley and the surrounding rural area of Zaire are heavily infected by AIDS. If the AIDS virus had been administered to children in Burundi in 1958 through polio vaccine immunizations, would not a much higher percentage of infected adults than 0.7% be observed in the rural areas? Yet it is the urban area in which the population is heavily (25 to 30%) infected with HIV, and it would seem that the spread in Africa occurs from urban areas to rural areas, not vice versa.
If one were to say that, since we gave polio vaccinations to the population in Leopoldville (now Kinshasa), capital of the former Belgian Congo, we could have started the AIDS epidemic there, we can point out that the same pool of virus material was used for the vaccination of children in L´opoldville and of children in Poland (9). Inasmuch as the prevalence of AIDS in Kinshasa is 25 to 30% and Poland has the lowest incidence of AIDS in Europe, one would have to undertake super-speculative acrobatics to incriminate the vaccine as the source of AIDS in Africa. Even the supposedly early cases of AIDS in Africa were clinically diagnosed several thousand kilometers away from the Kivu region (10). And the first proved case of AIDS was in a British sailor who died in 1959 and who showed symptoms of the disease throughout 1958, before any mass vaccination for polio was started (11).
After the original batch of the type II polio vaccine was produced in cotton rat brain (2), all other batches were produced in kidney tissue obtained from rhesus monkeys (Macaca mulatta) captured either in India of the Philippines (12). India's ban on the shipment of their monkeys outside India did not affect delivery from the Philippines. Curtis' speculation that we could have used in our production kidney tissue from other species of monkeys that might have harbored a simian immunodeficiency virus (SIV) or an HIV virus has no basis in fact. Importers of rhesus monkeys were carefully scrutinized by the U.S. government, and they could not supply tissue from animals coming from mixed sources. In addition, even if one speculates that green monkey tissue could somehow have been mixed up with rhesus monkey tissue, it has been shown that neither embryonic nor kidney tissue removed from SIV-infected African green monkeys contains SIV (13), (14).
Curtis attempts to make a case by saying "the Koprowski Congo preparation" may have been contaminated with an "unidentified cell-killing virus" found by Albert Sabin (15). First, Sabin did not test the Congo vaccine, but rather a seed lot of virus. Second, the same material in which he detected the presence of an unknown agent was retested in our laboratory and in two other labs, and no extraneous virus was found (16). Finally, it was reported at the Second International Conference on Live Polio Vaccines in 1960 that many monkey kidney vaccines infected with different strains of poliovirus, including strains developed by Sabin, contained "vaccuolating agents" cytopathic for tissue culture and contained foamy viruses. Such viruses were also isolated from polio noninfected cultures of both rhesus and cynomologous monkeys (17). I have presented cogent arguments (18) that the presence of these viruses does not disqualify the polio vaccine for worldwide use. I have, however, said that when a permanent culture of a human diploid cell strain free of all extraneous contaminants becomes available, it should be used for the production of poliovirus vaccine (14). The desire to replace the monkey kidney vaccine did not arise from concern about its safety. As a scientist, I felt that cells of human diploid culture kept permanently in a culture that could be kept at a limited passage level in vitro and thoroughly scrutinized for lack of extraneous contaminants should be recommended for vaccine production. Lots of vaccines produced in human diploid cells have been used for immunization in Switzerland (19) and in Croatia (20).
Among the 7,239,000 children vaccinated in Poland (21), 34,000 in Switzerland (22), and 1,500,000 (one-fourth of the entire population) in Croatia (23), careful follow-up observation has proved the vaccine to be safe and unassociated with any major illness. Even after the vaccination of 76,000 children in L´opoldville (9), it was possible to observe the vaccinees long enough to establish the protective effect of the vaccine at the time of an epidemic of poliomyelitis (24). Again, there was no doubt about the safety of the vaccine because there were no untoward reactions that could be attributed to an extraneous agent (24).
Curtis does not distinguish between lots of vaccines and seed lots of virus. There is no vaccine stored at the Wistar Institute, but there are a few vials of tissue culture supernatants available that may represent seed lots used for production of vaccines in the years 1957 to 1959. Curtis (3) suggests testing these samples "by multiple PCR [polymerase chain reaction] and other tests in independent labs by investigators of unquestioned integrity and stature outside the United States [italics mine] -- preferably in England and Switzerland." Any competent AIDS research lab could qualify to test these samples provided positive and negative controls are included in the assay. If the seed lot is found free of exogenous retroviruses (which is highly probable), contentious individuals could still argue that this does not represent what might have been present in the large lot of vaccine.
The argument for the safety of polio vaccination lies in the absence of any AIDS-related disease among the hundreds of millions of people vaccinated throughout the world; the fact that AIDS is rampant in subequatorial Africa can only be attributed to the polio vaccine by the wildest of lay speculation (25).
Tremendous efforts were made by scientists to save children from paralytic polio. The current anxiety among parents of children who have been or are going to be vaccinated against polio followed dissemination by the lay press of unproved theories of the origin of AIDS. This was unnecessary and harmful, particularly since the vaccine was tested thoroughly before any vaccination was done; the vaccine was and continues to be safe.
Jefferson Medical Center
Philadelphia, PA 19107-6799
(as given in the corrected version 11 September 1992, p. 1463).
7. Programme national de sero-prevalence de l'infection par le virus de l'immuno-deficience humaine (VIH) on Burundi. Juillet 1989-Fevrier 1990 (Ministre de la Sant´ Publique, Bujumbura, Republique du Burundi, 1990), p. 12.
25. Curtis states that he is not "the author of the theory" of the origin of AIDS. He says that the idea came to him from Blaine F. Elswood, whose paper written with R. B. Stricker "has been accepted by Research in Virology." This paper is still not published.