The origin of Aids

by Matt Ridley

Most scientists believe that Aids was "naturally" transferred from primates to human beings via a hunter who ate a chimpanzee. But a competing theory claims that Aids was caused in the 1950s when thousands of Africans were given a live polio vaccine derived from chimp kidneys. The stakes are getting higher.

The death of William Donald Hamilton on 7th March 2000 reverberated through biology like a thunderclap. Hamilton was probably the most famous evolutionary biologist in the world; a man whose bold theories had given birth to fruitful fields of inquiry not once but three times.

It was not the manner of his death which shook his colleagues so much. He died of malaria contracted in the Congo rain forest, where he had been looking for chimpanzee faeces-a tragic but not unsuitable death for a great naturalist whose curiosity about the animals of the tropical forest was an abiding theme of his life. What was disturbing was the reason why he was in the Congo in the first place. He was there in pursuit of an unfashionable, not to say cranky, theory: that Aids was caused by polio vaccination.

There was a general feeling, only muttered, that Bill might have gone a touch eccentric in endorsing this theory. Great scientists can go that way towards the end of their careers. (Linus Pauling believed that vitamin C could cure cancer. Fred Hoyle thought flu came from outer space. Alfred Russel Wallace became a spiritualist.) They seem to lose their scepticism. Some of Hamilton's biologist colleagues were therefore embarrassed by his conversion to an unfashionable conspiracy theory. Theories about Aids have a tendency to become conspiracy theories, blaming either the medical profession or the military. But Hamilton was not often wrong. And it now looks as if he may be proved posthumously right about this one, too.

A month after he died, a big American laboratory bowed to pressure which it had been resisting for eight years, and released for independent testing five samples of a polio vaccine held in a deep freeze. A meeting of the Royal Society, to discuss the theory, had already been arranged for this May by Hamilton and two Aids researchers. In the wake of his death this meeting assumed symbolic significance. Several prominent Aids researchers announced that they would boycott it in protest at the respectability it lent to the vaccine theory. The Royal Society postponed the conference until the autumn-to the anger of Hamilton's family. The stakes are getting higher.

Like all good conspiracy theories, the polio vaccine theory's originators are its worst enemies. Chief of these, Louis Pascal, wrote long and angry polemics from a New York address and refused to meet even his supporters. The theory was first aired in public in 1992 in a long article in Rolling Stone magazine-which did not encourage scientists to take it seriously.

Then it came to the attention of Edward Hooper, an unusually tenacious man. Hooper is British; he has spent much of his life in Africa, doing jobs ranging from storekeeper at a diamond mine to working for the BBC in Uganda. He was already writing a book about the origin and history of Aids, his second book about the disease. He had investigated several theories of origin and found most of them insubstantial or implausible. At first he thought the vaccine theory implausible, too. But gradually he found himself unable to dismiss it; worse, he found that the facts fitted it rather well. Worse still, when he asked those closest to the subject to disprove it to him, he was confronted not with contradictory evidence but with bluster, threats of legal action and angry denials.

You cannot pull the wool over Hooper's eyes. He checks everything. He digs out old colonial newspapers from the archives of Belgian libraries. He calls on the widows and children of long-dead scientists who might have had something to do with part of the story. He tracks down relatives of those who died of Aids early in the epidemic. So when he began to find glaring holes in the arguments put forward to deny the polio vaccine theory, his curiosity was piqued. For the next seven years he pursued the evidence, eventually writing an extraordinary and detailed account of the polio vaccine hypothesis and its rivals, entitled The River. In that book Hooper revealed many new facts about Aids and about what happened in the race to develop a polio vaccine in the 1950s. He came tantalisingly close to linking the two for certain.

But The River, although brilliantly written and carefully argued, was too thorough for this impatient age. It took the reader down every blind alley. It went into microscopic detail about long-forgotten laboratory procedures. It was studiously careful not to speculate beyond the established facts. Some readers found it hard to see the forest of theory for the trees of detail. What follows here is, for the impatient, the story that Hooper constructed.

Live vaccines are infectious viruses which have been rendered comparatively harmless. According to Hooper, a particular type of live polio vaccine called Chat may have been grown in the 1950s in cells derived from chimpanzee kidneys. Chimpanzees are the probable animal source of the Aids virus; live vaccines could have been contaminated if an infected animal was used. Chat was tested on more than 1m Africans in 1957-60, in the very areas where Aids subsequently became epidemic for the first time. Two other, less serious forms of Aids developed in parts of west Africa at about the same time, each epidemic closely associated with an area in which similar live polio vaccines may have been tested.

Stated thus, the theory seems purely circumstantial. It boils down to seven assertions, all of which must be tested to destruction.

First, chimpanzee kidney tissues will prove to have been used to grow Chat polio vaccine.

Second, those kidneys and the resulting vaccine will sometimes prove to have been contaminated with the chimpanzee SIV virus, simian Aids.

Third, the chimpanzee subspecies with the SIV closest to the main Aids virus (HIV-1, group M) will prove to be either the eastern subspecies of the common chimp or the bonobo. (Both kinds were kept at the camp in the Congo where the kidneys were taken.)

Fourth, no Aids or HIV case will prove to predate the vaccine trials.

Fifth, the first cases of HIV-1, group M, will prove to coincide in time and place with Chat vaccine trials in Congo and Burundi.

Sixth, the lesser outbreaks of HIV-1, caused by groups O and N, will prove to coincide with French vaccine trials in Gabon and Cameroon.

Finally, the epidemic of the less virulent HIV-2 will prove to be concentrated mostly in those parts of Guinea-Bissau where Portuguese vaccinations occurred in the 1960s.

The first assertion is about to be tested. Three different laboratories are being sent tiny droplets from a frozen sample of Chat polio vaccine which was kept in the Wistar Institute in Philadelphia, where the vaccine was developed. Hilary Koprowski joined the Wistar Institute as director in 1957, from Lederle Laboratories, bringing experimental polio vaccine strains with him. Just before the move, he visited the Congo, spent some days with a man who had grown live polio virus in chimp kidneys, and paid a visit to the Lindi camp, a facility near Stanleyville (now Kisangani) where wild chimps in large numbers were being held for medical experiments.

The test should prove whether chimpanzee kidney tissue was used to grow the vaccine-something which Koprowski has always denied, but without giving a convincing account of what other species of primate was being used. It is known that some chimp kidneys were sent to Philadelphia and to Belgium from Stanleyville during 1957, for medical experiments.

Exactly where the Chat vaccines used in Africa were prepared remains uncertain. Some undoubtedly came from the Wistar Institute; some came from Belgium, either from the Rega Institute in Leuven, or from a company called RIT. But some may have been made in the laboratoire médical in Stanleyville, where 400 chimpanzees were kept and killed between 1956 and 1958.

Even if chimpanzee tissue was used, the second assertion, that the vaccine was contaminated with SIV, will be difficult to prove or disprove. The five Wistar samples will be tested for virus; the results should be announced this summer. But many different batches of the vaccine were made, and most chimpanzees do not carry SIV, so not all batches would be contaminated. In a sample of 400 chimpanzees, though, it is probable that some were infected with SIV. In this respect, it is notable that there was an upsurge of infections by the bacterium Klebsiella among the chimpanzees kept at Lindi. Klebsiella is a normally harmless bacterium which turns virulent in Aids patients, especially in Africa, and in monkeys and apes which suffer from simian Aids.

That monkey viruses can contaminate live vaccines is not in doubt. The SV40 virus, which can cause cancer, was discovered in 1960 in rhesus monkey kidneys. It was not the first, but rather the 40th contaminating virus to be discovered in monkeys alone. By the time it was screened out, SV40 had already contaminated tens of millions of doses of live polio vaccine. It was known to cause cancer in mice. Yet the medical profession, seeing no upturn in the incidence of cancer, breathed a sigh of relief. Only in the 1990s, thanks to determined work by an Italian scientist called Michele Carbone, has it emerged that SV40 is strongly implicated in the upsurge of pleural mesothelioma or asbestosis in recent decades, and possibly other cancers as well, especially brain tumours. Asbestos seems to be much more carcinogenic in people infected with SV40.

Even if the chimps were infected with SIV, could it have contaminated kidney tissue culture? Earlier this year, at a conference on Aids in San Francisco, scientists from New York announced that "the kidney appears to be a previously unrecognised reservoir for HIV-1 infection" in human beings. Even the most carefully prepared culture of kidney cells can contain small numbers of lymphocytes, or white blood cells, which are the natural target cells for SIV. The process of vaccine production is now so improved that no contaminating viruses could survive preparation of the culture, but that may not have been the case in the 1950s. The surviving protocols for the preparation of the Chat vaccine are too vague for us to be sure.

The third assertion-that the chimps from the area near Lindi carry the virus most similar to the main human Aids virus-is the one Bill Hamilton was hoping to test during his fatal expedition. At present, only four samples of SIV have been isolated from chimpanzees, and three of these are from the "wrong," or western subspecies. They resemble HIV-1, groups O and N, more than they resemble HIV-1, group M.

The remaining sample of chimp SIV is from a chimp of the eastern subspecies-the same subspecies that was kept at Lindi. Its SIV is also like groups O and N, which seems to undermine the polio theory. But Hooper points out that you cannot rely on one sample alone; the chimp in question, named Noah, spent a long time in captivity, first in the Congo and then in Belgium, and could have been cross-infected with an SIV from another primate. Moreover, there was a different species of ape, the bonobo or pygmy chimp, at Lindi, and no SIVs have yet been found in this sister species. Unless Hamilton's samples reveal some more SIVs, direct from the Congo, this third assertion will remain unresolved either way.

The fourth assertion, that no case of Aids predates the vaccine trials, was rejected as false in 1992 by a committee appointed by the Wistar Institute. At the time there was one well-known case of Aids from Manchester in 1959, surviving samples from which seemed to contain HIV. Presumably, if the man had died in 1959, he must have been ill for many years before that. Hooper discovered that this man had indeed fallen ill as early as 1956, while in the Royal Navy. However, the samples were re-tested and found to be HIV-negative: it was a simple case of laboratory contamination by a modern HIV virus.

With the demise of the Manchester patient, the oldest cases of definite Aids in the west are isolated case histories dating from the 1960s and early 1970s-a Norwegian and his family, two Belgians, two Germans and an American. A thorough search of the medical literature turned up some other potential cases-even a possible epidemic near Czech uranium mines- but close investigation revealed that none of these cases would today have been described as Aids.

All but one of the early Aids cases in the west have plausible African connections-even the two German cases occurred in people living close to where the Zaïre (now Congo) football team and its supporters were during the World Cup of 1974. However, the earliest American case, a rug-addicted 16-year-old mother in New Jersey, who gave birth to a HIV-positive child in 1973 or 1974, did not have African connections. Intriguingly, she would have been a baby at the time that the first trials of the suspect batch of Chat vaccine were carried out, in 1957, in a New Jersey women's prison. The babies born there were vaccinated.

As for HIV itself, the oldest positive test now consists of a blood sample taken from an unknown African man in Kinshasa (then Leopoldville) in 1959 in unrecorded circumstances. Since there was a Chat vaccination trial in the city at the same time, it is possible that this sample was taken in a post-vaccination follow-up. So it cannot be ruled out that this man had contracted his HIV from polio vaccine a few weeks before. So far, therefore, the fourth assertion looks remarkably strong. Human Aids dates from the same years during which live polio vaccines were tested.

The fifth assertion, that the Chat vaccine trials coincide in time and place with the first outbreaks of HIV-1, group M, is the most circumstantial-but also the most impressive. The key years are 1957-60, when the first batches of Chat vaccine were given. They were used in Stanleyville, in parts of northeast Congo, in some towns in other parts of Congo and in the capital, Leopoldville. They were also used in Poland and Switzerland. The European tests apparently resulted in no cases of Aids, but in both places the numbers of people given the relevant batches of vaccine were small.

In Africa there is good correlation with early cases of HIV. Moreover, there are other parts of Congo, such as Kasai province, where no vaccination was carried out and no early HIV or Aids was recorded. The correlation is not perfect, but some of the anecdotes are suggestive. A Belgian cartographer and his Congolese wife left Africa in 1968 to retire to Belgium, where they were both found to have very early cases of Aids. They had been living in Kikwit in 1959, when Chat vaccinations were being given to Europeans living in the town. Nearly all of the samples of HIV-1 from Africa in 1980 or before came from places where the Chat vaccine was used.

The most important correlation is with the largest of all the Chat trials-in the Ruzizi Valley, in eastern Congo, and the western part of Burundi, along the shore of Lake Tanganyika. There, 215,000 people were vaccinated with two batches of Chat in early 1958. There, in exactly the towns and villages which were fed with the second batch, extraordinarily high rates of HIV prevalence were found in 1980-even before Aids was recognised, and long before such high rates were found even in African cities, let alone in rural areas. In Rumonge, a small fishing port in western Burundi where Chat had been given to most of the children in 1958, 12 per cent of the population was HIV-positive in 1980, four times higher than the incidence in the city of Kinshasa. This western Burundi correlation is striking. The first place in the world which received a mass vaccination with live polio vaccine is the first place in the world to suffer a mass HIV epidemic. Correlation is not necessarily causation, but it can be very suggestive.

The sixth assertion-that a similar contamination occurred in French colonies in Africa-is less well grounded. It is known that Pierre Lépine of the Pasteur Institute developed a live polio vaccine in 1957, and he later referred to "experiments" in the face of a polio epidemic. It is also known that a doctor in Mitzic, in northern Gabon, vaccinated more than 2,000 people with "the Lépine vaccine of the Pasteur Institute in Paris" in response to a polio epidemic in November 1957. These seem to be descriptions of the same event, in which case a live vaccine was probably used at Mitzic. Since Mitzic lies at the heart of the area in which HIV-1, groups O and N, have been found, again the correlation is good; but there are no more details. For the Lépine vaccine to be the cause of this rarer HIV strain, it would have had to be prepared in chimpanzee tissues. No information about this is available.

The early cases of Aids in a Norwegian family were of the O strain. This fits with the fact that the father, the first case, had served in the navy and caught gonorrhea while on a voyage which took him to Cameroon in 1961. That appears to be where he caught HIV, too.

The seventh assertion is that the "other Aids epidemic" of HIV-2 in west Africa resulted from a vaccination campaign in the Portuguese colony of Guinea-Bissau. Here again Hooper has found no direct evidence, but he has found suggestive hints. There are memories of vaccination by the Portuguese in the 1960s, and good correlations between those parts of Guinea-Bissau which were controlled by the Portuguese in 1969 and the areas where HIV-2 is now common. The areas which were then controlled by rebels are now much less affected by the epidemic.

Guinea-Bissau is much the worst country affected by HIV-2. Neighbouring Liberia and Sierra Leone are comparatively free of the virus, especially in rural areas. Only in parts of Senegal, where Guinea-Bissan refugees settled in the 1960s, and in Côte d'Ivoire, are there concentrations of HIV-2 which even approach the prevalence of those in Guinea-Bissau.

HIV-2, a much less virulent pathogen, does not resemble the chimp SIV, but rather the sooty mangabey SIV. It is therefore surprising to find that sooty mangabeys do not live in Guinea-Bissau; indeed, they have been extinct there for much of this century. This casts doubt on the hypothesis that the epidemic started more naturally, with a hunter who cut himself while skinning a monkey, because sooty mangabeys are numerous in Liberia and Sierra Leone, where they are both kept as pets and hunted for food. Sooty mangabeys have been extensively used for medical experiments, however, and were exported from west Africa in large numbers for this purpose in the appropriate years. But there remains no direct evidence that they were used by the Portuguese to produce vaccines.

Most scientists favour the theory that Aids reached humanity via a cut, scratch or mouth sore on a hunter who ate a chimpanzee (or, in west Africa, a mangabey).

Asked to explain why four separate strains of SIV made the jump to human beings at about the same time, they reply that Aids is an ancient disease which has sporadically erupted in human beings but has previously died out, rather as the Ebola virus does. What was different, in the 1950s, was the sudden development of urbanisation, sexual promiscuity, cheap long-distance travel, warfare, and medical records-all of which combined to turn a hitherto sporadic infection into a global pandemic.

Hooper documents a serious defect in this reasoning. Central and western Africa were turbulent long before the 1950s. The history of the region is dominated by the effects of the slave trade, with mass population movements, frequent warfare and abundant sexual promiscuity. During colonial times, especially in the Congo, huge numbers of people were moved to mining areas as forced labour, where they lived in dense slums and prostitution was rife. Yet no Aids outbreaks occurred in such places-places in which causes of death were well recorded. Not a single HIV virus reached the Americas with the 10m Africans taken there by the slave trade, although other blood-borne African viruses did.

Moreover, the correlation between human beings eating primates and Aids is poor. As we have noted, Guinea-Bissau has sooty mangabey Aids, but no sooty mangabeys. In northern Congo the pygmies eat primates, but have no HIV. The "natural transfer" theory simply does not deserve the confidence placed in it by most scientists. The burden of proof has been shifted.

One piece of evidence seems to support natural transfer, and that is the genetic differences among different HIV strains. The ten or so different forms of HIV-1, group M, are all sufficiently different to assume that, with normal rates of evolutionary change in the genes of the virus, they must have shared a common ancestor in the 1940s or even earlier. Hooper has no problem with this. Either evolutionary change has been unusually rapid, as it often is in a new host species, or, just as plausibly, the ten groups represent ten different chimpanzees in the cages at Lindi, with slightly different viruses. Those chimps were caught from many different wild troops.

Hooper's book has certainly stirred things up. Act-Up, an Aids campaigning group, has tried to organise a protest outside the Wistar Institute, where the originator of the Chat polio vaccine, Hilary Koprowski, shelters behind his lawyers. But Hooper refuses to endorse such demonstrations, believing that the problem is not a matter of blame but of historical research. The contamination, if it happened, was inadvertent.

Nevertheless, the implications are terrible. Even without the Aids link, what Hooper has found is appalling. A live vaccine, whose safety had not been properly tested, was given to hundreds of thousands of Africans, few of whom could benefit from it (many were immune to polio). That there was the most serious outbreak of polio in Kinshasa's history shortly after the vaccinations there, suggests that the vaccine might have reverted to virulence, and set off the polio epidemic. British doctors warned at the time that this was a danger, so ignorance was no excuse.

Worse, the scientists who carried out this useless experiment never followed it up to see if there had been any safety problems. Nor did they record exactly how they made the vaccine. Nor, of course, did they have much compunction about capturing hundreds of young wild chimpanzees (often killing the parents in the process), keeping them in dreadful conditions, and removing their organs-perhaps even before killing them. (One black technician in Kisangani, interviewed by Hooper and Hamilton, remembers vivisection of chimps which were paralysed but apparently conscious.) Even by the standards of the 1950s, this was pretty low. Little wonder that Hooper has found it so difficult to persuade the scientists involved to tell him what they got up to.

Yet, as he points out, if Aids did not derive from a contaminated vaccine, they have nothing to fear from the truth. If they did not use chimpanzees, then surely they could produce evidence of what they did use. If they know that the vaccine could not have been contaminated, then surely they can produce the protocols and do the experiments to show it. It is no longer enough simply to say that the hypothesis is too speculative and does not deserve to be tested. In terms of establishing cause and effect, it is now rather less speculative than the theory that salt causes high blood pressure.

Hooper's opponents argue that knowing where Aids originated is irrelevant to stopping its progress today. It makes no difference to the fight against Aids whether it was an accident or an act of God. Indeed, by discrediting vaccination, the Hooper argument might actually make the task of fighting Aids more difficult, should an Aids vaccine ever be developed. Hooper retorts that the truth matters.