by Pepin
What is the evidence in support of Haiti having been infected from the US or vice versa?
In retrospect, the first probable cases of AIDS in Haiti were recognised in 1978–9. This retrospective chronology is based essentially on diagnoses of Kaposi’s sarcoma (KS), a cancer which had rarely been diagnosed in Haiti prior to 1979. This is the form of skin cancer developed in the character played by Tom Hanks in Philadelphia, for which he received an Oscar. AIDS-associated Kaposi’s sarcoma involves internal organs but patients also display multiple skin lesions which can be easily biopsied, provided the patient has access to a hospital equipped to do histopathological examinations. AIDS-associated KS has to be distinguished from ‘endemic Kaposi’s sarcoma’. While the former is associated with a profound immunosuppression and a high mortality, the latter is an indolent cancer, compatible with prolonged survival, which has long been recognised as endemic in parts of Africa, including the Belgian Congo, AEF and Uganda. Among eighteen cases of KS diagnosed in Haiti or among Haitians in the US in 1979–81, most died within six months of diagnosis: clearly, this was not the ‘endemic’ form of the disease. There was a male preponderance.54–61
Similarly, cases of probable AIDS with KS among gay men were identified retrospectively in the US in 1978–9. Such temporal coincidence suggests that whichever country got HIV first, the other one got it not long after. The median incubation period from acquiring HIV until developing AIDS is about ten years, but can be shorter in some patients. These observations suggest that HIV was introduced into both Haiti and the United States at the end of the 1960s or in the early 1970s.62–65
However, it does not imply that HIV was introduced into both countries at exactly the same time, for several reasons. KS is not the best marker for retrospectively recognising the emergence of AIDS in a population. Kaposi’s sarcoma is a cancer caused by another sexually transmitted virus, human herpesvirus 8. During the early years of the American epidemic, Kaposi’s sarcoma as an AIDS-defining illness was much more common in homosexuals than in other risk groups, presumably because human herpesvirus 8 is transmitted better during homosexual than heterosexual intercourse. KS was seen in 21% of homosexuals with AIDS, but in only 6% of male heterosexuals and 1% of haemophiliac men who developed AIDS. Once infected, for some reason males in general are intrinsically more susceptible to the cancer-causing effect of the virus than females (fifty years ago, in central and East Africa, endemic KS was five to thirty-three times more common in men than women). By analogy, changes in the incidence of KS in Haiti probably reflected the introduction of HIV into its homosexual/bisexual community rather than among individuals who acquired HIV through other modes.59–61,66
Access to a diagnosis of Kaposi’s sarcoma was infinitely better in the US than in Haiti. In the US, it would have been very unlikely for somebody with KS not to have a skin biopsy with histopathological interpretation and registration of the case in a cancer registry. In Haiti, it is plausible that some of the early cases of KS were missed. It is interesting to note that a case of AIDS (without KS) was diagnosed in Montreal in 1978, in a Haitian who went to Canada for medical treatment. Since only a tiny proportion of Haitian patients would have had the contacts and resources to travel abroad to seek medical treatment, there must have been earlier cases of AIDS who died quietly in Haiti without a diagnosis. The most frequent HIV-associated opportunistic infection in Haiti would have been tuberculosis, a disease that was already so common in the impoverished island, long before HIV emerged, that any change in its incidence or clinical pattern would have taken years before it was noticed.67
As in Africa, scientists tried to locate archival samples of serum. Out of 191 Haitian adults from a rural area tested for dengue fever in 1977–9, none was HIV-positive. Molecular biologists came to the rescue of historians and provided estimates of the respective chronology of the two apparently concomitant epidemics in Haiti and the US. A first study using a molecular clock estimated that the founder of the B subtype in the US originated in 1967 (confidence interval: 1960–71). In the phylogenetic tree, the seven B subtype sequences from Haiti ‘branched off’ earlier than the other B subtype sequences, which suggested that HIV in Haiti antedated its introduction into the US.68–69
More precise measures were generated when researchers recovered HIV-1 sequences from archival specimens collected at a Miami hospital in 1982–3 from Haitian AIDS patients who had recently emigrated to the US and had presumably been infected with HIV-1 while in Haiti. These sequences were compared to isolates from the US and other countries, all of which were HIV-1 group M subtype B. If HIV-1 had arrived in Haiti first, non-Haitian subtype B strains would be expected to be phylogenetically nested within an older and more extensive range of Haitian genetic variations, with Haitian lineages branching off closest to the ancestor. This is exactly what the analysis showed. The probability that subtype B emerged in the US prior to Haiti was estimated at less than one in a thousand.70
Analyses supported the hypothesis of a single epidemiologically successful introduction of subtype B from central Africa to Haiti, from where it was re-exported to the US. The time of the most recent common ancestor of subtype B in Haiti was estimated to be 1966 (confidence interval: 1962–70) while the most recent common ancestor for the US epidemic was estimated at 1969 (confidence interval: 1966–72). In other words, HIV was introduced into Haiti around 1966, and from there it moved to the US around 1969, give or take a few years. This was consistent with another study which dated the founder of the US type B epidemic at 1968.70–71
These findings represent the best available data concerning the chronology of the introduction of HIV-1 into the Americas. They were contested by Haitian researchers, and one can understand that the wounds of the anti-Haitian stigmatisation of twenty-five years ago have left permanent scars, especially considering that these were superimposed on centuries of domination and exploitation of Haitians by white westerners. However, their arguments were refuted by the authors’ reply published in the same journal.72,73
There is now little doubt that HIV-1 subtype B was exported from central Africa to Haiti around 1966, from where it was re-exported to the US a few years later. Among the 4,500 Haitians who worked in the Congo, one of them acquired HIV-1, probably through heterosexual intercourse, and later initiated a chain of transmission upon returning to the Caribbean island, during vacations or at the end of his contract. As in any population of 4,500 adults, there must have been a small minority who were sexually promiscuous and bought sex once in a while. The same behaviours that facilitated acquisition of HIV-1 within the heart of Africa must have contributed to its early spread in Haiti, the returning technical assistant infecting one or more Haitian women, perhaps a sex worker.
Molecular biology and phylogenetics aside, we can be relatively certain of the number of Haitians who introduced HIV-1 into the Americas: a single individual. That is because subtype B, the exclusive subtype present among Haitians and Americans in the early stage of the epidemic, is very uncommon in central Africa where it represents less than 0.5% of all HIV-1 strains that circulate. Thus, it is virtually impossible statistically that more than one Haitian working in the Congo got infected with the same subtype in the early 1960s and brought it back home. This is an extraordinary example of what evolutionary biologists describe as a founder event.
But why was the introduction of HIV-1 into Haiti so epidemiologically successful rather than just another dead end infection, as with the Norwegian sailor and the Belgian expatriates infected in the Congo in the 1960s? How was it possible for a virus imported in 1966, by only one individual, to infect 8% of mothers attending an under-five clinic in the Cité Soleil slum of Port-au-Prince merely a decade and a half later, bearing in mind that such an expansion required more than fifty years in Léopoldville/Kinshasa? There must have been a very effective amplification mechanism early on, but was it sexual or parenteral? This will be the topic of the next chapter.74
12 The blood trade
I
n this chapter, we will examine the possibility that, during the early stage of the Haitian epidemic, a commercial enterprise in Port-au-Prince exponentially and parenterally amplified the number of HIV-1-infected individuals and allowed the virus to thrive. More generally, we will review the role of the blood trade in the globalisation of HIV-1. But first, we need to understand how viruses can be transmitted, not only from donor to recipient, but also from one donor to another during the handling required to prepare certain blood products. The word ‘donor’ is somewhat misleading here because we are talking mostly about people paid for their ‘donations’.
Blood is made of cells (red blood cells, white blood cells and platelets) and plasma, its liquid component. Plasma is made of water and proteins: antibodies, clotting factors and albumin. When a donation is made, the various components are separated to maximise their use. Patients with anaemia or acute blood loss need only receive the red blood cells, those with a low platelet count will be given the platelets and so on. Plasma is highly valuable as it contains many proteins. Therapeutic use of plasma started during WWII as an expander of intravascular volume, to increase quickly blood pressure in patients with serious bleeding. Subsequently, other uses of plasma components were developed, which required the selective processing of specific proteins: albumin (to expand intravascular volume or to patients with low albumin levels), coagulation factors (haemophilia or other coagulation disorders) and immunoglobulins (patients with immune deficiencies or to protect travellers against hepatitis A).
Plasma was also used for the production of the early generation of hepatitis B vaccines, made from the chemical inactivation of the virus present in blood and the purification of its surface antigen. Sources of hepatitis B-positive donors included gay men and prisoners in developed countries, and the general population of Third World nations, where up to 15% of adults chronically carry HBV in their blood. At least 30 million doses of such crude vaccines were administered before being replaced with genetically engineered vaccines.
As the amount of the specific proteins of interest is small in each individual donor’s plasma, commercial plasma derivatives must be prepared by pooling plasma from several donors. In particular, the fabrication of coagulation factor concentrates required the pooling of plasma from thousands of donors, such that an infectious agent present in a single donor could make the entire pool infectious and be transmitted to many recipients. Prior to their discovery, transmission of HIV-1 and HCV from coagulation factor concentrates infected thousands of haemophiliacs worldwide, resulting in the tragic death of a large portion of this population. Since then, the infectious risk from coagulation factors has been drastically reduced through the development of sensitive screening assays for donors and better methods for eliminating viruses.
The risk of transmitting HIV was much lower, indeed probably near zero, when the plasma was processed to prepare albumin, because ethanol was used in the fractionation, after which the product was pasteurised by heating. And fortunately, no transmission through immunoglobulins has ever been reported, even if many batches contained antibodies against HIV: again, the ethanol fractionation process inactivated HIV. The early hepatitis B vaccines were not incriminated in the transmission of HIV either; presumably, the methods used to inactivate HBV were effective against HIV as well.1–2
In the late 1960s and 1970s, before HIV and HCV were known, the demand for plasma-derived products had escalated rapidly. There was not enough excess plasma from volunteer whole blood donors and paid donors had to be recruited. In order to get more plasma from these paid donors, a technology called ‘plasmapheresis’ was developed: whole blood was taken from the donor, the plasma quickly separated from the blood cells and the cells re-infused in the donor along with replacement fluids. Thus the donor did not become anaemic and could sell plasma repeatedly, not just twice a year as with donors of whole blood. However, before and even after the infectious risks were understood, viruses could be transmitted not only to the ultimate recipients of the blood products but also between donors participating in plasmapheresis. This required only one breakdown in some component of the process, for instance the re-use of pieces of plastic tubing that had been designed for single use. If a donor with unrecognised HIV infection entered the process and if some precaution was disregarded, this person could infect subsequent plasma donors whose blood was processed by the same machine on the same or following days. In settings where paid donors repeatedly sold their plasma week after week, this would increase the number and proportion of HIV-infected individuals among those selling plasma, further enhancing the risk for the other donors. This vicious circle would result in exponential propagation of HIV between donors, arguably the most effective method for HIV transmission.
The infectious risks for donors in plasmapheresis centres had been known for some time before the HIV epidemic. In 1973, donors at a South Carolina commercial plasmapheresis centre contracted the hepatitis A virus, a microbe which remains present in the bloodstream for only a short period of time. This was caused by the pooling of plasma from multiple donors during its extraction from the cells, allowing the reflux of pooled plasma into the bags of red cells re-infused in the donors. In 1977–8, four outbreaks involved plasma donors who developed ‘non-A non-B hepatitis’ (later renamed hepatitis C after its aetiological agent was discovered) in plasmapheresis centres in Austria, Germany and Poland, apparently from contamination with plasma of the plastic bags used for re-infusing red blood cells. In the US, paid donors were often recruited in prisons, a substantial percentage of whose populations were previous or current drug addicts at high risk of being infected with HBV or HCV: a chain of transmission could easily be initiated.3–5
The first well-documented epidemic of HIV-1 among paid plasma donors occurred in a poor suburb of Mexico City where, in 1986, 281 donors were found to be HIV-infected, especially those that sold plasma ten or more times each month. Re-utilisation of blood collecting material was blamed. At the time there were thirteen plasmapheresis centres in the country, mainly in Mexico City and in states near the Texan border. Most donors were young men living in the peri-urban shanty towns. They could sell their blood as often as every two to three days. By the time that the sale of plasma was prohibited nationwide in 1987, 7% of 9,100 paid donors were HIV-infected. In one of the plasmapheresis centres, HIV prevalence increased from 6% in June to 54% in November 1986. Other outbreaks of HIV transmission among paid plasma donors were reported in Valencia, Spain and Pune, India. In the latter city, among commercial plasma donors, HIV prevalence was 0% in November 1987 but 78% seven months later, illustrating the exponential transmission of HIV through unhygienic plasma collection practices.6–11
These outbreaks, although tragic in their own right, were dwarfed by what happened in China in the early 1990s, several years after the risk of HIV transmission via plasmapheresis was understood, and a decade after the transmission of HCV had been documented in the same Chinese centres. In rural areas, poor farmers were recruited by ‘plasma pimps’, to sell plasma to increase their meagre income. They received $6 per donation, which could be repeated twice a month in theory, more often if donors attended more than one collection centre. There were several hundred plasma collection stations set up by blood product companies. In the most-heavily affected provinces of Henan, Anhui, Shanxi, Hubei, Hebei, Shandong and Jilin, approximately 250,000 paid donors (a quarter of a million!) acquired HIV.12–14
In several plasma collection centres, blood from multiple ABO-matched donors (who were not screened for HIV) was combined for ‘more efficient’ large volume plasma separation, and then the pooled cell fraction was returned to the donors, along with any infectious agent that had been present in the blood of any of the donors at a given session. The re-use of needles and tubing also facilitated transmission. In some regions, between 9 and 17% of plasma donors became HIV-infected while up to 28% were infected with HCV. It is remarkable that such a high HIV prevalence was reached despite most donors reporting fewer t
han ten donations per year. Among the small number who sold plasma more than twenty times per year, half became HIV-infected.12,15
What do all these stories have in common? Poor people looking for a quick source of income and willing to sell their blood repeatedly. Profit-driven blood collection centres where a small number of entrepreneurs try to make as much money as possible by cutting costs, re-using needles, syringes and tubings, while being unaware of or not caring about the risk of transmitting blood-borne viruses. A lucrative market for these blood products, either locally or internationally. Finally, a ‘patient zero’ who introduces the pathogen.
The vampire of the Caribbean
Now back to the Caribbean, where the potential for a quick profit in the blood trade had been exploited as well. In Port-au-Prince, a large plasmapheresis centre operated from May 1971 to November 1972 under the name Hemo-Caribbean. This was a joint venture between Joseph Gorinstein, a Miami businessman and stockbroker, a few other American investors and a well-known Haitian politician, Luckner Cambronne.
Luckner Cambronne was born in 1929, the son of a poor Protestant preacher. Starting out as a bank teller, he eventually found a job in the entourage of François Duvalier (Papa Doc), the country doctor elected president in 1957. Initially just a messenger, he then became a bagman. Duvalier liked him and Cambronne quickly rose to become the regime’s chief extortionist. He held various ministries (public works, customs, etc.), all of which provided ample opportunity for corruption. His speciality was to intimidate businessmen into making large ‘donations’, and those who refused had a much shorter life expectancy. Ostensibly, these funds were to be used to rebuild a slum or pave a road but most of it ended up in Duvalier’s and Cambronne’s bank accounts. This was also the main destination for the funds (deducted from the pay of civil servants) allocated to building a new city, Duvalierville. Only a few bungalows were erected, far from the promised Caribbean Brasilia.16
