The Origins of AIDS

by Pepin

  The fact that in parts of Cameroun and Gabon up to 50% of some birth cohorts were infected iatrogenically with HCV suggests that if SIVcpz/HIV-1 was ever introduced into the group of patients receiving injections in a given healthcare facility, the amplification must have been substantial. So from one SIVcpz/HIV-1-infected patient, there could have been hundreds after a few years. Such amplification was probably limited geographically to the patients receiving medical care in the same hospital, dispensary or disease control mobile clinic as the first SIVcpz-infected patient. However, once the number of HIV-1-infected individuals had increased to a few hundred, the probability of at least one of them moving to a city, infecting a free woman and initiating a chain of sexual transmission increased proportionally. From a long shot this became an unavoidable certainty.

  Such an initial phase of parenteral amplification was necessary for the emergence of HIV-2 in West Africa, a virus whose sexual transmission is rather ineffective but which still managed to infect tens of thousands of individuals, as summarised in Chapter 10. HIV-2 slowly faded away when the opportunities for its parenteral transmission decreased: the number of newly infected subjects ultimately became less than the number of long-infected individuals who died of AIDS or some other disease. This contrasts with HIV-1 group M which, after benefiting from the parenteral amplification, also proved its superior ability to infect lymphocytes and other cells, a characteristic that would facilitate its sexual transmission. Part of the epidemiological success story of HIV-1 group M was thus probably biological and intrinsic to this virus. HIV-1 group N did not benefit from such parenteral amplification, because it crossed species in a more recent era when IV drugs had been replaced by oral meds for the treatment of many tropical diseases, with the result that it remained limited to a very small number of infected individuals.

  While the initial phase of parenteral amplification of HIV-1 group M must have occurred in an area inhabited by P.t. troglodytes, the next phase of sexual dissemination followed migration and trading routes. From a rural area, HIV-1 moved into the nascent cities of central Africa. On a purely geographic basis, the most likely candidates for such urban spread were Yaoundé, Douala, Libreville, Bangui, Brazzaville and Léopoldville. However, given the very low HIV-1 prevalence measured in Cameroonian and Gabonese cities in the mid-1980s, among the general adult population but also sex workers (Chapter 13), the early sexual propagation probably did not occur there. The clear preponderance in Cameroon and Gabon of CRF02A_G, a recombinant which could not have emerged in the early years of the epidemic, further rules out these locations as the site of the initial spread. We cannot rule out Bangui, where prevalence was already a significant 4.4% in 1986. It is unfortunate that no archival serum samples from Bangui or southern Central African Republic were ever found in the local Institut Pasteur or that if such work was done it was not published. The lower genetic diversity of HIV-1 subtypes in Bangui compared to Kinshasa and Brazzaville (Chapter 1) suggests that the virus has been circulating in the latter area for a longer period of time than on the banks of the Oubangui. Furthermore, only a small part of the Central African Republic is inhabited by P.t. troglodytes and the genetic diversity of HIV-1 is similar to that in Chad, a country without any P.t. troglodytes. These two countries’ intermediate genetic diversity of HIV-1 isolates probably represents an ancient import of subtypes which had already been differentiating somewhere else.5–6

  Several arguments suggest that the Léopoldville/Brazzaville urban area was central in the dissemination of HIV-1 (Chapters 1 and 13). First, the extraordinary diversity of HIV-1 in samples obtained from Kinshasa and Brazzaville, where all known subtypes and many recombinants are present. Second, the two oldest samples containing HIV-1 obtained in 1959–60 both came from Léopoldville. Third, the retrospective description of five clear-cut cases of AIDS that corresponded to infections probably acquired in the DRC in the late 1960s or early 1970s. Fourth, the presence of HIV-1 antibodies in samples obtained in 1970 from women representing the general adult population of Kinshasa (mothers bringing their children to an under-five clinic), even if at a low prevalence of 0.25%. Fifth is the finding that in 1980, among a similar sample of women, prevalence was already at 3.0%. Sixth, the surveys conducted in representative samples of the adult population of various central African cities in the mid-1980s showed low prevalences in Gabon and Cameroon but rather high prevalences (4–8%) in Kinshasa and Brazzaville. Finally, the status of the Léopoldville/Brazzaville conurbation, for a long time the commercial and administrative heart of central Africa attracting thousands of migrants each year into an ever larger melting pot of all ethnic groups.7–15

  What was the geographical origin of the first cases of HIV-1 imported into Léopoldville/Brazzaville? There is a remote possibility that the source was from within the Belgian Congo. The part of the Bas-Congo region north of the river was inhabited by P.t. troglodytes, and there were excellent road and rail communications between Boma, Matadi and Léopoldville. However, this area is small, with limited P.t. troglodytes populations, and we do not know whether local communities of chimps are infected with SIVcpz. Far more likely, the virus managed to reach Brazzaville first, with its subsequent implantation in Léopoldville just across the river. As reviewed in Chapter 5, Brazzaville became the administrative centre of the AEF federation, the terminus for navigation on the Congo and the departure point of the train to Pointe-Noire and the Atlantic coast. These factors implied constant movements of Africans and Europeans between the capital city and the territories of Oubangui-Chari, Moyen-Congo and Gabon. Initially concentrated within Moyen-Congo, these population movements reached further inland and on a massive scale when forced labour was brought in for the construction of the Congo–Océan railway in the 1920s. Furthermore, the part of south-east Cameroun with the highest prevalence of SIVcpz among its chimps, and whose strains of SIVcpz are closest to HIV-1 group M, has rivers, the Dja and the Ngoko, that drain into the Congo system through the Sangha, which also drains part of southern Oubangui-Chari inhabited by P.t. troglodytes. For these populations, trading was easier towards the Congo than to Yaoundé and Douala, and ever since the German colonisation, steamboats would regularly make the journey between south-east Cameroon and Brazzaville/Léopoldville.

  Once HIV-1 reached Brazzaville, it would not take long for it to move into Léopoldville. Some traders from the AEF hinterland visited both cities on the same or successive journeys, depending on where they thought they could fetch the highest prices for their goods. Alternatively, there might have been a little local transmission within Brazzaville before the virus was introduced into Léopoldville. There was a ferry as well as smaller boats departing every half hour for the twenty-minute journey. Before independence, commerce was brisk between the two cities, as traders, many of them women, managed to find a price differential for some goods that would make the short trip profitable. And after independence in 1960, when the economic situation deteriorated in the DRC, trading if anything increased as agricultural products from the Bas-Congo could be sold at higher prices in Brazzaville, where smart traders could then use their CFA francs to buy goods that Léopoldville was short of, guaranteeing a hefty profit. Free women also moved back and forth between the two capitals while other migrants settled on the other side of the river for longer periods in search of a better life (Chapters 5 and 6).16

  Once HIV-1 was present in both cities, the conditions in Léopoldville were more favourable to its successful sexual propagation than in Brazzaville. Early in colonial history, a gross gender imbalance was created in Léopoldville by the Belgian colonists, on a scale far worse than across the river, as we have examined in Chapter 5. For a few decades, Léopoldville was an urban work camp where women and children were unwelcome. This was fertile ground for prostitution to develop on a large scale, and develop it did. Even when colonial policies were softened after WWII, Léopoldville was such a booming town that it constantly attracted a flux of migrants from the adjacent rural areas and young
men would come first, perpetuating the imbalance. On both sides of the trade, prostitution involved mainly the unmarried and within this subgroup as late as the 1950s there were more than five men for each woman in Léopoldville.

  For the first few decades of its presence in Léopoldville, the dissemination of HIV-1 was slow and limited. According to mathematical models, for a long time there was something like 100 infected individuals in the city (Chapter 3). The type of sex trade that existed during the colonial era corresponded to what is currently referred to as concomitant partnerships or semi-prostitution rather than hard-core prostitution (Chapter 6). A free woman would have a few regular clients, perhaps three on average, to whom she provided a variety of services, not just sex. If that woman got infected with HIV-1, she could only transmit the virus to one of her other steady clients, who might eventually infect another free woman. This setting was good enough for the persistence of the infection, possibly a very slow growth of the infected population up to a few hundred, but nothing like the exponential transmission of HIV-1 that was to be documented in Kenya in the early 1980s.

  It is likely that some of the initial transmission of HIV-1 to free women in Léopoldville occurred iatrogenically and not just sexually. For a long time, starting in the early 1930s, free women were forced to attend the Dispensaire Antivénérien in Léo-Est for regular STD screening (Chapter 9). Those with a positive syphilis serology (many of whom carried such antibodies not because of syphilis but because they had had yaws in their childhood) were treated with injectable drugs, most often administered IV. In 1953, more than 150,000 injections were administered just in this one institution, which treated up to 1,000 patients each day. Documentary evidence reveals that syringes and needles were not sterilised but only rinsed between patients, with the result that many cases of iatrogenic hepatitis B acquired in this STD clinic were recognised by a clinician at the main hospital of Léopoldville, even in a setting where only a small minority of adults was susceptible to infection with HBV. This situation created an extraordinary opportunity for the spread of HIV-1: the women infected iatrogenically were semi-prostitutes, who could in turn transmit the infection, now sexually, to some of their regular clients. And thus a perfect storm developed.17

  More widespread sexual transmission became possible when the face of prostitution in Léopoldville was dramatically altered around 1960–1. The political chaos and civil war in parts of the Congo brought hundreds of thousands of internal refugees into the capital, resulting in massive unemployment and poverty (Chapter 11). As documented by several observers, high-risk prostitution appeared, with sex workers providing sexual services to a few men every day, potentially more than 1,000 per year, in downmarket brothels which were little more than glorified shacks. For a number of years, it remained a concentrated epidemic, with a higher prevalence among prostitutes than among their clients. As long as the gender imbalance persisted, that is until the early 1970s, there were relatively few opportunities for dissemination outside this initial core group, because for many men it was not easy to find a female partner to marry or with whom they could at least sustain a romantic relationship. The transition from a concentrated to a generalised epidemic occurred between 1970 and 1980, a period during which HIV prevalence among mothers at a well-baby clinic in Kinshasa increased from 0.25 to 3.0%. If these prevalences were representative of the whole adult population of Kinshasa, the number of HIV-1-infected individuals in the city rose from about 1,400 in 1970 to some 36,000 ten years later. That is why in the mid- to late 1970s physicians at the Mama Yemo and at the university hospitals started seeing cases of what would be later recognised retrospectively as AIDS.13,14,18–21

  Once the number of HIV-infected persons in Léopoldville expanded, it was unavoidable that it would spread outside the capital, which was the political and economic centre of a large country with endless movement of traders, bureaucrats and all kinds of economic migrants between the various regions of the Belgian Congo, the DRC and later Zaire. HIV-1 eventually reached Kigali, Rwanda, where it found extremely favourable conditions for its sexual transmission in this small city with a gross surplus of uncircumcised men and a thriving sex industry (Chapter 13). A similar process, albeit with a different founder strain, happened in Bujumbura, Burundi. In other countries of central Africa, HIV-1 arrived at the same time or even earlier, but as conditions were less propitious its spread was slower. From central Africa, HIV then progressively moved into other regions of the continent. Southern Africa was infected by the extension of subtype C infection from the Katanga province of the DRC to Zambia, and then further south to Zimbabwe, Malawi and eventually South Africa. East Africa was infected via Kisangani while West Africa was infected by the extension of the CRF02A_G recombinant northwards along the coast, from Gabon and Cameroon to Nigeria, Ivory Coast and so on.

  At some point between 1960 and 1966, among the 4,500 teachers dispatched to Congo-Léopoldville, a single Haitian technical assistant was infected with HIV-1 group M subtype B (Chapter 11). Around 1966, this person went back to Haiti and stayed long enough to start a local chain of sexual transmission. There, this rare subtype of HIV-1 had to be amplified exponentially early on, otherwise it would have been impossible for a virus introduced in 1966 to infect 8% of women in Port-au-Prince sixteen years later, reaching a level seen in Léopoldville–Kinshasa only several decades after its introduction.22–23

  Did this amplification occur sexually, within the small homosexual/bisexual community of male Haitians selling sex to American tourists, as suggested by some? I doubt it very much, for a simple reason: American gay tourists did not stay in Port-au-Prince or elsewhere in Haiti for long enough. For exponential amplification to occur through the homosexual route, as documented later in San Francisco, New York and many other locations, a first HIV-infected person needs to transmit the virus to one or two individuals, then each of these second-generation cases needs to infect a few more, and each of these third-generation cases to infect a few more, and so on, with a short interval (three or four months) between each cycle of transmission. Most American tourists who acquired HIV-1 subtype B from a Haitian male prostitute probably went back to the US within a couple of weeks, before they developed the high viraemia and high infectiousness which is characteristic of primary HIV infection and which drives the sexual amplification of HIV. Thus, these tourists had little opportunity to infect a second Haitian male prostitute for such a vicious circle to be initiated but a much greater chance of infecting other American gay men back home. Some American gay tourists undoubtedly infected Haitian male sex workers, or acquired HIV-1 from these same prostitutes, and some homosexual prostitutes infected each other, but it does not seem plausible that this caused a massive spread of the virus within the small Haitian bisexual community severe enough for the infection to spill over quickly into the heterosexual population.24–26

  Admittedly, this part of the story remains unproven, but there are good reasons to believe that the Hemo-Caribbean plasmapheresis centre (Chapter 12) could have been the perfect breeding ground for a quick increase in the number of HIV-infected Haitians, to a level where sexual transmission would then inevitably but more slowly allow for its further dissemination. Extremely rapid transmission of HIV-1 was documented among individuals selling their blood in commercial plasmapheresis centres in China, Mexico and India. Once the virus is introduced among paid plasma donors, up to three-fourths can get infected within a year. It seems unlikely that the procedures that should have prevented the transmission of blood-borne viruses were more stringent in Port-au-Prince than in all these other places, quite the contrary. Haitian plasma sellers were even poorer than their counterparts in other countries and they had to put up with whatever was done at the Hemo-Caribbean clinic. For this disaster to have happened, we need just one HIV-1-infected person to have entered the cohort of plasma sellers; within the next year, from a handful of HIV-1-infected persons on the island, there could have been several hundred. At this stage, the number of infect
ed persons in Haiti had reached the critical mass which enabled it to disseminate successfully among sex workers, male and female, and then into the general adult population.

  If indeed this happened, then the export of the plasma to the US, where it was bought and processed by large pharmaceutical companies, could have allowed for some spread of HIV-1 subtype B into the Americas and Western Europe. Once a shipment of plasma entered the stock of a plasma broker, it could be sold and resold in several countries on both sides of the Atlantic within a short period of time (Chapter 12). We do not know whether all of the plasma sent to the US or Europe by Hemo-Caribbean was processed into albumin and immunoglobulins (with no risk of HIV transmission), or if some lots were used to prepare factor VIII cryoprecipitates, in which case this business could have contributed to the spread of the virus. Either through this route or via American gay tourists, HIV-1 was introduced into the US. It was already present, albeit at a low prevalence, among drug addicts in the mid-1970s, a population within which HIV spread quickly: by 1979, one third of addicts in New York were infected.27–28