The Strange Journey Of HIV From Kinshasa To New York

The simian immunodeficiency virus carried by Pan troglodytes troglodytes in the rainforests of south-eastern Cameroon is not, in itself, a remarkable pathogen. Chimpanzees have lived with it for a very long time. What turned a chimpanzee infection into a global human catastrophe is the geography of empire, the chemistry of cheap injectable medicine, and the unhappy coincidence of several modernising states all reaching for syringes at the same moment.

Molecular clock work pins the chimp-to-human jump for HIV-1 group M, the pandemic strain, somewhere around the turn of the twentieth century. The most cited estimate places it near 1908, with a confidence interval running from the late 1880s to the early 1920s. The crossing almost certainly happened more than once. Group M took over the world. Groups N, O and P, all sourced from the same broad region, mostly did not. The difference is not virological mystery. It is plumbing, transport, and demography.

A hunter in the Sangha basin in 1910 had the virus inside him and almost no chance of starting a pandemic. A clerk in Léopoldville in 1925 had everything he needed.

Léopoldville Built An Epidemic

Léopoldville, renamed Kinshasa in 1966, is where the molecular evidence converges. Nuno Faria and colleagues, writing in Science in 2014, used a dataset of 800 archival HIV-1 sequences to date the ancestral group M lineage to the city, with an emergence window of roughly 1920. Every major HIV-1 group M subtype radiates outward from there. The phylogenetic tree is not subtle. It points at one river port in colonial Belgian Congo and asks the rest of the world to explain itself.

Three things made Léopoldville a virological accelerator.

1. The first was the Belgian colonial transport system, which connected the city via the Congo river and Kasai river to a hinterland of mining and rubber towns. Sleeping sickness control, smallpox campaigns, treatment for yaws and syphilis, and routine clinic work all required injections. Syringes were scarce. Needles were reused, sometimes thousands of times per unit, sometimes with cursory flame sterilisation, sometimes with none. Jacques Pépin, the Sherbrooke epidemiologist whose Origins of AIDS did more than any single text to reconstruct the early amplification, estimates millions of parenteral procedures across Equatorial Africa in the first half of the century. A blood-borne virus could not have asked for better infrastructure.
2. The second factor was the sex ratio. Colonial labour policy in the Belgian Congo deliberately produced a city of unattached men. Women were a minority. Commercial sex was, predictably, ubiquitous. By the 1950s, ulcerative sexually transmitted infections including chancroid and untreated syphilis were endemic in Léopoldville, and ulcerative STIs are one of the most efficient amplifiers of HIV transmission known.
3. The third was independence. The Belgian withdrawal in 1960 left a state without the trained professionals required to run it. The new government recruited from francophone elsewhere. A great deal of that elsewhere happened to be Haiti.

Haiti Became The Certain Uncertainty

The orthodox account runs like so: several thousand Haitian doctors, teachers, agronomists and civil servants worked in Congo between roughly 1960 and 1966. Some returned home. One or more of them brought subtype B with them. The virus then circulated quietly in Haiti for a few years before stepping onto a flight to New York or Miami around 1969, founding the North American epidemic.

The argument leans heavily on a 2007 paper in PNAS by Michael Worobey and Marlea Gilbert, which examined HIV-1 subtype B sequences from five Haitian immigrants diagnosed in the early 1980s . It concluded a single migration of subtype B from Haiti to the United States, dated by molecular clock to about 1969, with a 95% confidence interval reaching back to 1966.

The paper presented its case carefully. The textbooks did not.

By the time the finding reached general circulation, the qualifier had been worn off, and Haiti had become a confident waypoint rather than a plausible candidate.

The Haitian leg is doing a lot of work for a fairly small evidence base. Five sequences from patients diagnosed years after the event is not, by the standards of modern phylogenetics, an overwhelming sample. The dating depends on assumed mutation rates, on the representativeness of the Haitian-immigrant cohort, and on a model preferring a single founder event to several. The single-founder model is not a discovery. It is a parsimony assumption, and parsimony is a useful heuristic, not a finding.

There is also the awkward question of what the Haitians actually did when they returned. Haiti is not exactly the most welcoming place for LGBT types. Nor was sodomy particularly prevalent in the past.

In 2017, gay "marriage" was simply banned by law.

A returning teacher infected in 1962 would, in most reconstructions, have remained asymptomatic for around a decade. To seed an epidemic in Port-au-Prince and then export it to North America by 1969 requires either an unusually busy sexual network or some other mechanism of transmission running in parallel.

As it happens, there was one.

Hemo-Caribbean And The Vampire

In the early 1970s, Port-au-Prince ran a commercial plasmapheresis operation called Hemo-Caribbean, controlled by Luckner Cambronne, the Duvalier regime's interior minister. Cambronne was known locally as the Vampire of the Caribbean for reasons which did not require subtlety.

Hemo-Caribbean paid Haitian donors a small sum per litre, extracted plasma in conditions later described by visiting journalists as squalid, and exported up to 6,000 litres of plasma per month, principally to the United States and to European pharmaceutical firms. The operation ran from roughly 1971 until international pressure shut it down in late 1972.

A 1972 New York Times report described the facility in unimpressed terms. American firms including Cutter Laboratories and Armour Pharmaceutical were named as buyers. The product was pooled, fractionated into clotting factors and immunoglobulins, and distributed widely.

What this means for HIV epidemiology is uncomfortable.

The Haitian-origin story of an American epidemic depends on plausible routes of viral export. Sex tourism by American gay men to Port-au-Prince exists in the literature, often as the dominant explanation, sometimes accompanied by a hint of editorial relish.

Some of it is utterly foul.

The scene is a gay bath house in San Francisco in 1982:

“When the moaning stopped, the young man rolled over on his back for a cigarette. Gaetan Dugas reached up for the lights, turning up the rheostat slowly so his partner's eyes would have time to adjust. He then made a point of eyeing the purple lesions on his chest. "Gay cancer," he said, almost as if he were talking to himself. "Maybe you'll get it, too."

Plasma export is mentioned less often.

It is, on the face of it, the more efficient mechanism. A single infected donor, pooled with others, fractionated and distributed to thousands of haemophiliacs and surgical patients, would have produced a transmission pattern of a particular form, including the early appearance of AIDS in haemophiliacs which is exactly what was observed in the United States by 1982.

Pépin argues the plasma trade is probably how subtype B reached North America, and it is a more uncomfortable story than the one which settled into textbooks. It implicates regulators rather than tourists. And it connects the with infected blood scandal in the UK of the same era where the NHS was warned not to import supplies from the US because of the hepatitis problem.

It's a lot more believable than sodomy outbreaks on an island defined by religiosity.

It also means the Haitian leg, to whatever extent it existed, may have been less a cultural artefact and more a procurement failure by American pharmaceutical industry, with the Duvalier regime supplying the raw material.

Phylogenetics Picks A Founder

Worobey returned to the question in 2016 with a substantially better dataset. The Nature paper, drawing on archival serum from the 1978 and 1979 New York and San Francisco hepatitis B vaccine cohorts, recovered eight near-complete HIV-1 genomes from before AIDS was clinically recognised.

The genetic diversity present in those samples was striking.

Subtype B in the United States by 1978 already carried the signatures of multiple years of cryptic circulation. The paper put the most probable US entry around 1970 or 1971, with New York as the early hub and onward spread to San Francisco around 1972.

This is consistent with a Haitian intermediate.

It is also consistent with several other arrangements which Worobey did not rule out, including direct African-to-US transmission alongside or in place of the Haitian route, or multiple seedings from a Caribbean reservoir which itself drew from several African sources. The 2016 dataset improves the dating. It does not adjudicate the route.

A simplified timeline pulled from the major phylogenetic and historical work looks roughly as follows.

The interesting feature of this table is the eleven-year gap between US entry and clinical recognition. Whatever the route through the Caribbean, the virus was in New York for the better part of a decade before American medicine noticed.

The Hepatitis B Vaccine Connection

The hepatitis B vaccine cohorts run by Wolf Szmuness at the New York Blood Center, and the parallel San Francisco cohort enrolled from 1978 to 1980, were assembled for the most prosaic of reasons. Hepatitis B was rampant in sodomy networks. A vaccine candidate needed a high-incidence population for efficacy testing. Roughly 1,083 men entered the New York trial. Around 6,875 entered the San Francisco cohort. Blood was drawn at intervals of three to six months and frozen.

The trial achieved its primary aim. Heptavax-B worked. It was approved in 1981. It was also, by accident, the most comprehensive biological archive of the early American HIV epidemic anyone could have assembled deliberately, and nobody did assemble it deliberately, because nobody knew what they were assembling.

When the stored sera were later tested, the results were brutal.

• The New York cohort showed 6.6% HIV positivity by 1978 or 1979 and 43.7% by early 1984.
• The San Francisco cohort moved from 4.5% in 1978 to 67.3% in 1984 and 73.1% in 1985.
• Six of the first ten reported San Francisco AIDS cases in 1981 turned out, on later cross-reference, to have been members of the hepatitis B cohort.

A particular nutjob conspiracy theory has lived inside this set of facts for forty years. It says the plasma-derived hepatitis B vaccine introduced HIV to American gay men.

The theory survives on the surface appearance of the data, which is suspicious enough to keep noticing, and dies on the contents, which include the inconvenient detail of pre-vaccination HIV positivity in the trial population.

The cohort was selected from a network already saturated. The vaccine arrived after the virus, not before it. The trial produced the archive. It did not produce the epidemic.

Patient Zero Was A Filing Error

The figure of Gaétan Dugas, the Air Canada flight attendant from Quebec identified for years as Patient Zero, deserves a brief and unflattering visit in any honest account of the period.

Dugas appeared in a 1984 CDC cluster study as Patient O, with O standing for Out-of-California, a designation referring to his residence outside Los Angeles rather than any priority in the epidemic. The letter became a number in subsequent retelling. Randy Shilts, writing And the Band Played On, accepted and amplified the framing. A man who was, on the genetic evidence, an entirely typical mid-stage case became a folk villain.

The 2016 Worobey paper sequenced Dugas's archived blood. His virus sat in the middle of the American subtype B tree, not at its root. He was a passenger, not a captain.

The Patient Zero story was, throughout, a journalistic construction sitting on a clerical confusion, and the fact of its endurance into the 2010s says more about the appetites of the press than about the epidemiology of HIV.

Belgian Governors And Haitian Dictators

The standard account of HIV's journey from Kinshasa to New York is correct in its broad outline. What we now understand is firmer picture with an empirical basis.

1. A chimpanzee virus crossed into humans in early twentieth-century Equatorial Africa.
2. Colonial medical infrastructure amplified it in Léopoldville.
3. It spread through African urban networks for several decades.
4. It reached the Americas around 1970.
5. By the time medicine in San Francisco and New York noticed, it had been spreading for a decade and was already entrenched.

What the account does less well is the Caribbean middle.

The Haitian intermediate is treated with a confidence the underlying evidence does not really support. The phylogenetic case for a Haitian staging post is suggestive rather than conclusive. The plasma trade through Hemo-Caribbean, which is the most parsimonious mechanism for moving the virus from Port-au-Prince to American patients in the relevant window, gets a fraction of the attention given to gay sex tourism, despite being better documented and more epidemiologically tidy.

The story preferred by the textbooks places the moral weight on the appetites of American homosexuals. The story preferred by the documentary record places at least some of it on the procurement practices of American pharmaceutical companies and the export operations of a Haitian dictatorship.

There is a pattern here worth noticing.

Pandemic origin stories tend to drift, over time, toward whichever villain the period finds congenial. In the 1980s, the dominant story was a moral one about sexual behaviour, and the Haitian intermediate fit it by supplying an exotic conduit. The actual record contains exotic conduits, but it also contains industrial ones, and the industrial ones are less satisfying to write about.

The other persistent smoothing concerns colonial responsibility.

The Belgian Congo's role in amplification was not incidental. The pandemic strain of HIV-1 became the pandemic strain because of decisions taken by a European colonial administration about how to deliver injectable medicine cheaply, how to staff a city with single men, and how to regulate prostitution.

Those decisions did not cause the chimp-to-human jump. They did cause the human-to-everyone jump. Histories of HIV which begin in the 1980s with American hospital wards tend to omit this. The omission is convenient. It is also wrong.

A clean reading of the evidence, then, looks like this:

1. HIV-1 group M became a pandemic because Belgian colonial medical infrastructure converted a rare zoonotic event into a sustained urban epidemic.
2. It became a North American epidemic because, somewhere in the late 1960s or early 1970s, it crossed the Atlantic, probably via Haiti, and probably with at least some contribution from a plasma export trade which American pharmaceutical industry preferred not to dwell on after the fact.
3. It became visible in 1981 because by then a critical mass of immunocompromised young men had reached the same hospitals at the same time, and Pneumocystis pneumonia in the absence of an obvious cause is the sort of clinical pattern even an inattentive system eventually notices.
4. All of this relates to the NHS infected blood scandal.

Everything else is commentary, and most of the commentary is downstream of whichever villain the writer arrived with.