Letter to Science

Tom Curtis, 1992 letter to Science (not published)

[This letter was later published as an appendix in Julian Cribb, The White Death (Sydney: Angus and Robertson, 1996), pp. 258-262.]

4425 McKinney Street
Houston, Texas 77023-1127
(713) 921-3646
September 30, 1992

Ms. Christine Gilbert
Letters Editor
Science Magazine
1333 H Street, N.W.
Washington, D.C. 20005

To the Editor:

As a journalist, I’m glad to be a catalyst for scientific discussion of the theory that the AIDS epidemic was spawned by a simian immunodeficiency virus (SIV) that may have contaminated early live oral polio vaccines tested on about a quarter million Central Africans more than 30 years ago.

Now the developer of that vaccine has weighed in (21 August 1992). But many points that Hilary Koprowski invokes to rebut my report in Rolling Stone (1) are contradicted by or unsupported by his references. Such arguments should not divert scientists from urging testing of his vaccines (2).

Koprowski addresses my presentation of the theory rather than the proponents’ because Blaine F. Elswood and Raphael B. Stricker, whose paper (3) provided my initial road map, and Louis Pascal, who was the first to publish on this topic (4), haven’t been permitted to express their ideas in a scientific periodical. (The Journal of Medical Ethics, however, called Pascal’s hypothesis “important and thoroughly argued” and said it “ought to be taken seriously by workers in the AIDS field” (5).) My response reflects help from all three theorists, notably Pascal.

This theory proposes that an uncommon event occurred sometime between 1957 and early 1960: a rare HIV-1-like SIV contaminated as little as a single batch of Koprowski’s vaccine tested in Burundi, Rwanda, and the former Belgian Congo (now Zaire). Perhaps just a fraction of one percent of those receiving the vaccine became infected with that agent, and those infected people infected others, who infected others, to form the current worldwide epidemic of AIDS. The posited infection could have originated with laboratory contamination of the monkey kidneys used to grow polio virus (as from a dirty scalpel previously used to dissect a chimp with an HIV-1-like virus) or from using a monkey with the atypical HIV-1-like virus. Intriguingly, scientists recently reported finding a single African monkey among 500 sampled that tests positive for such a virus (6,7).

Koprowski correctly observes that most Congo vaccinations occurred in northwestern Burundi, not in the immediately adjacent Kivu district of Zaire as I had it. And he is right that the very high rates of HIV-1 positives that I cited in Kivu were later scaled back (I was unaware of the subsequent report when I wrote for Rolling Stone). Yet both the Kivu district and the area where these vaccinations occurred are within the area of the map I cited showing the region of highest AIDS incidence in Africa (8). In fact, the area of greatest vaccination lies near the center of this map. To brand such data “completely wrong” because the location was less than 40 miles off and because the incidence cited has been reduced obscures the larger reality.

Contrary to Koprowski’s implication, the earliest African cases of AIDS did not arise “several thousand kilometers away” from where his vaccines were dispensed. Three of the four that can be localized appear extremely close. Not a single case cited in his reference (9) is demonstrably more than 1000 km from a vaccination site — and the first was in Leopoldville (Kinshasa), where about 75,000 children were inoculated (10). The three remaining cases could range from 0 to no more than 1300 km from one of the two sites.

Koprowski asks: If the AIDS virus were in a vaccine administered to rural children in 1958, wouldn’t one see a much higher percentage of infected adults today in these rural areas, rather than in the cities, where rates are higher? Not necessarily. Only small numbers might become infected initially. This is because orally ingested HIV is presumed destroyed by digestive acids. SIV-contaminated live vaccine sprayed into Africans’ mouths, as Koprowski’s vaccine was (10), might infect only those with mouth sores, bleeding gums, infected tonsils, impaired mucosal immunity, or those who inhaled the mist or got it in their bloodstream some other way. Arguably, those infected as infants in 1958 would have died long before passing the virus on, well before testing began in the mid-1980s. Those infected closer to puberty who migrated to cities would have had the best chance to spread the disease. The highest rates of HIV-infection are in cities because that is where people have the most sexual contacts.

Rates of infection in Poland, where the same vaccine was used, are too low to be consistent with the theory, Koprowski contends. This argument fails to consider both of the most important variables in determining the number of infections to be expected today — the number vaccinated and the rate of spread after vaccination. Somewhat over 300,000 were given the vaccine in Africa (10,11); somewhat more than 3000 were given the same allegedly contaminated African vaccine in Poland (12,13). Thus, initial infections in Poland should have been 1/100 of those in Africa.

The growth rate also would be expected to be different. Africa has one of the highest growth rates for AIDS largely because the disease spreads heterosexually there with great efficiency; it spreads heterosexually much less efficiently in Poland and other European nations. If 50 people infected by the African virus survived to pass the virus on, and if the number infected doubled 17 times in the 34 years since inoculation, this would yield the 6.5 million Central African cases estimated today. By that standard, the 3000 or so Polish vaccinations would yield zero to one infections.

Assuming that the doubling time in Poland was twice Africa’s (thanks to lower rates of promiscuity), a single original infection in Poland in 1958 could have produced about 192 HIV-1 infections now. As of April 30, Poland reported 103 cases of frank AIDS, suggesting perhaps 660 infections. Early European cases were imported from Central Africa and occurred in France and Belgium, which had colonies there, and from the U.S. Poland lacks strong ties to Central Africa, to European counties with high rates of infection, and to the U.S. Thus, its low numbers fit the theory.

Seven million Polish children did not get the same vaccine preparations used in Africa, as Koprowski implied. He has said elsewhere that the Polish vaccines were made by Wyeth Laboratories (1), not at the Wistar Institute, which produced the African vaccines. Moreover, the 3000 in Poland inoculated early on with the African vaccine got it before Albert Sabin informed Koprowski of the contaminating virus he found in it (14). This allegedly contaminated vaccine likewise wasn’t used in Switzerland or Croatia, places Koprowski cites as evidence the vaccine was uninfected.

The case of the British sailor who died of AIDS in 1959 refutes the theory, Koprowski suggests, because his symptoms started before the vaccine was administered. But it is unclear whether his AIDS came from HIV-1 or HIV-2; and he may have travelled to Africa after Koprowski’s vaccine was first dispensed there in 1957 and 1958 (15). Further, the seaman’s classic AIDS symptoms weren’t manifested until December 1958 (16), not throughout 1958 as Koprowski says. By April 1958, more than 200,000 Central Africans had been vaccinated (11). While rare, rapid deterioration and death after contracting HIV-1 does happen (17).

Koprowski’s reference does not support his claim that only kidneys of rhesus monkeys captured in India or the Philippines were used to grow the Congo vaccines. He wrote this paper more than four years after he embarked on the Congo campaign and after Wyeth Laboratories had produced 7 million doses of vaccines for use in Poland (18). The context makes plain that he described then-current practice, not the manufacture of his Congo vaccines, when he said polio vaccine is grown “from freshly harvested kidneys from monkeys brought to the U.S. either from India or the Philippines.” To suggest that this verifies that Congo vaccines were made with “rhesus monkeys (Macaca mulatta),” and that rhesus monkeys from the Philippines were unaffected by the Indian ban on monkey exports, compounds the error: Rhesus monkeys do not come from the Philippines (19). Earlier Koprowski told me that the kidneys were already removed when received at Wistar (1); at that time he was unsure whether rhesus monkeys or African green monkeys — both of which can be infected with SIV — had provided his raw material.

A related argument, that green monkey kidney tissue from SIV-infected monkeys doesn’t contain SIV, does not really speak to the issue. It’s undisputed that primary monkey kidney culture, the medium in which polio vaccine is made, contains mononuclear cells such as lymphocytes and macrophages that may be infected with SIV.

Live oral polio vaccines are made by adding weakened poliovirus strains cultured in primary monkey kidneys, known as seed lots, to larger numbers of fresh monkey kidneys to make vaccine lots. Koprowski suggests that when I referred to Sabin’s paper reporting that he found a contaminating virus in Koprowski’s Congo vaccine, I should have said this allegedly was in a seed lot. The issue may be important for two reasons: If Sabin’s disputed finding was correct, it shows that the Congo vaccine contained at least one extraneous, pathogenic virus. Were this agent SIV and in a tested vaccine lot, then only a single monkey kidney and a single Congo batch of vaccine need have been contaminated.

Sabin’s referenced paper said his “tests on the large lot of Koprowski’s type 1 Chat vaccine [emphasis added] used in the Belgian Congo trials…revealed the presence of an unidentified, non-poliomyelitis cytopathogenic virus.” (14) Koprowski’s cited letter described “tests on the large Chat pool mentioned by Sabin” (20) [emphasis added]. Koprowski said his lab and the two others which found no extraneous virus did not perform the same test Sabin did because they lacked Sabin’s especially potent testing material.

After interviewing Koprowski, I certainly thought that samples of the Congo vaccine lots were stored at the Wistar Institute. If this is wrong, as he asserts, and if the fewer than 100 samples of poliovirus material recently found in freezers there and described by one authority as vaccine samples (2) include only seed lots, as Koprowski suggests, it is imperative that Koprowski and any others who know say where actual vaccine samples are stored.

The contention that “careful follow-up observation” proved the Congo vaccine to have been safe conflicts with what Koprowski has told me: No long-term follow-ups were possible because of the civil war in the Congo beginning in 1960 following the polio trials (1). To use a source written in 1963 in the midst of national chaos to prove that there were “no untoward reactions that could be attributed to an extraneous agent” begs the question — especially when one is discussing a slow virus that would have been introduced from 1957 through 1960 and has a latency period of ten years.

If, as Koprowski says, in 1961 his “desire to replace the monkey kidney vaccine did not arise from concern about its safety,” this is at odds with what he wrote at the time. While arguing that the polio vaccine was safe, Koprowski acknowledged that it would be “more difficult to justify scientifically a stand that nothing should be done about the host cells in which polioviruses are grown (18).” He added: “Not only has the existence in monkey tissue of the dreaded B virus (which is definitely pathogenic for man) been known for some time, but it is clear that the next batch of killed monkeys may contain more ‘virus surprises'”.

Among the many “virus surprises” found to have contaminated primary monkey kidney cultures, medical science discovered, developed tests for, and eliminated two especially troubling ones — SV40 and SIV — only after the vaccines potentially containing these agents had been administered to hundreds of millions of people. Lederle Laboratories division of American Cyanamid, the sole U.S. maker of oral polio vaccines since 1976, continues to produce polio vaccine using kidneys of African green monkeys, the host for SIV (a virus which was not discovered until 1984). While Lederle now tests for SIV and for some other monkey viruses (21), there is no way to test for as yet unknown “virus surprises.”

The federal government’s top AIDS researcher Anthony Fauci, Nobel laureate Frederick C. Robbins (who shared the prize for the discovery of how to grow polio virus in tissue culture), and eminent virologist and Polio Hall of Famer Joseph Melnick, without endorsing any theory, said last March that all available old polio stocks should be tested for the presence of SIV, HIV and related retroviruses using the most sophisticated tests available (21, 22).

Unfortunately, so far nothing of the sort has happened. Testing old polio stocks for HIV-1 and related retroviruses is something we owe to those with AIDS, to ourselves to preclude (if possible) future “virus surprises” from using monkey tissue, and to the subjects of future experimental vaccine trials in Africa and elsewhere. What we learn may well change how we make vaccines and prompt us to ban future simian-to-human organ transplants.

Sincerely yours,

Tom Curtis


1. T. Curtis, Rolling Stone No. 626, p. 54 (19 March 1992).

2. T. Curtis, P. Manson, Houston Post, p. 1 (July 17, 1992).

3. B. Elswood, R. Stricker, Polio Vaccines and the Origin of AIDS (unpublished manuscript) (1992).

4. L. Pascal, What Happens When Science Goes Bad, Science and Technology Analysis Working Paper No. 9, Department of Science and Technology Studies, University of Wollongong, Locked Bag 8844, South Coast Mail Centre 2521, Australia (December 1991).

5. R. Gillion, J. of Med. Ethics 18 p. 3, (1992)

6. G. Lecatsas, J. Alexander, Lancet 339:1427 (6 June 1992).

7. T. Curtis, Houston Post, p. 1, (2 August 1992).

8. D.B. Hrdy, Rev. of Inf. Dis., 9 (6):1109, Nov.-Dec. 1987

9. J. Sommet et al, Scand. J. Infect. Dis., 19 511 (1987)

10. A. Lebrun et al, Bull. WHO 22 203 (1960).

11.. G. Courtois et al, Brit. Med. J., p. 187 (26 July 1958).

12. F. Przmesmycki, et. al., Am. J. Hygiene, 71 275 (1960)

13. F. Przmesmycki, et al, Second Int’l Conf. on Live Polio Vaccines, p. 522 (1960).

14. A. Sabin, Brit. Med. J., 14 Mar 1959, p. 663. 15. D. Concar, Nature, 346, p. 95 (12 July 1990).

16. G. Williams et al, Lancet ii P. 951 (1960)

17. D. Ho, personal communication (August 1992).

18. H. Koprowski, J. Am. Med. Assoc. 178, 1151

19. J. Wolfheim, Primates of the World, U. of Washington Press, Seattle p. 490 (1983).

20. H. Koprowski, Brit. Med. J., p. 1349 (23 May 1959).

21. T. Curtis, Houston Post, p. 1, (22 March 1992)

22. T. Curtis, P. Manson, Houston Post, p. 1 (March 26, 1992).