Review of Tinderbox by Craig Timberg and Daniel Halperin
Review of Tinderbox by Craig Timberg and Daniel Halperin [Penguin USA; 2012]
This is powder puff reporting, as a ten-minute glance at the forward matter and the origins chapters (the Prologue and chapters 2 to 8 inclusive) quickly reveals. In the Acknowledgements, the authors thank several of the usual suspects, notably Beatrice Hahn and Michael Worobey, for "historical, cultural and epidemiological information" and "vital perspectives". Perhaps it is not surprising that these are the very same scientists who are most feted (embarrassingly so, in the view of many) in the text of the
This is the book that Worobey and Hahn themselves might have produced, were they writers.
Tinderbox doesn't even bother with the OPV theory, other than to claim that it has been disproved. And even then they get it wrong. On page 66 it is stated that "Edward Hooper and others seeking to explain the growth of HIV after independence have pointed to the proliferation of mass inoculation campaigns in the final years of colonial rule". But what I point to is a unique oral polio vaccination campaign that started the epidemic, not to colonial inoculation campaigns that boosted an already existing epidemic. As the authors of Tinderbox should know, oral vaccines are given by mouth and do not involve "inoculation". It is others, such as Preston Marx, who favour the theory of contaminated needles providing a boost to the spread of HIV. By conflating these two theories, they show that they are either ignorant on a very basic level about what people like Marx and I are saying, or else that they are deliberately attempting to spread misinformation. Let us take the first, less sinister option. In that case this is a sloppy mistake,
of a type that typifies a sloppy book.
No arguments that challenge the bushmeat theory are examined. (This is in contrast to my own book, "The River", in which I devoted an early chapter to examining different theories of origin, and explaining in some detail why most of them didn't work.)
When it comes to dating the epidemic, the problem of recombination is not even mentioned, other than a strange reference to "what even scientists call, with a smirk, 'virus sex'." (Actually, I think that scientists sometimes refer to 'viral sex', but the sentence that follows suggests that the authors actually don't realise that viral sex, or recombination, is different from mutation.)
Timberg and Halperin baldly state that the first chimp-to-human transfer of SIV occurred "two decades either side of 1900", with the best bet being 1908. Note that 1908 is 13 years earlier than Pepin's account, which (based on the same sources) proposes 1921 as the date of transfer. What is really striking is that the transfer date actually doesn't seem to matter that much, as long as it's before the OPV trials in the late 1950s.
P>As for the place: the authors swallow the Hahn and Worobey origins story whole, and baldly state that the first transfer of chimp SIV to humans took place in southern Cameroon. Our fearless reporters don't actually twig that there is no hard evidence to support this.
They say that Bill Hamilton "gushed" praise for my work in his foreword to The River. They don't point out that this very private man supported my work whole-heartedly from the first day we met (in 1993) to the last time we spoke (in January 2000, just before his second departure for the Congo, this time with Michael Worobey). By contrast, Worobey knew Bill Hamilton for less than six months, and spent just over 3 weeks with him in the Congo. During that time Worobey scratched his thumb in the jungle, but didn't return to Kisangani to get it treated until the last moment. Worobey once told me in interview (slightly proudly, I thought) that the Kisangani doctors had informed him that he had been hours away from losing his whole hand (and perhaps his life) to gangrene. And shortly after that Bill caught cerebral malaria, an attack that must surely have weakened him substantially, and perhaps contributed to his final demise from a duodenal haemorrhage shortly after his return to the UK. If we are talking about "gushing", then it is fair to point out that Timberg and Halperin describe this trip to the Congo in unmistakably gushing and heroic terms. In my opinion it sounds like a lousy expedition, one that demonstrated a lack of sound leadership and consistently poor judgement that endangered lives.
The first time I met Michael Worobey was while Bill was lying in a coma in University College Hospital in London, and within seconds his coolness made it quite clear that although Bill was dying, Worobey saw me as a rival for the great man's affections. Two years later, although I found it hard to like Worobey, I did my level best to trust him. On three occasions I drove the 200 miles to and from Oxford to discuss the possibility of collaborating with him on a study of ancient tissue samples. That idea fell though because in order to facilitate his writing of a grant proposal, he wanted me to provide specific details that (as I had already warned him) I was not prepared to divulge at that stage. After that he failed to answer a series of several emails I sent him over a five-month period, and it became apparent that we would not be able to work together. For two years he had assured me that he had doubts about how AIDS had started, and that he could see clearly that the factor of recombination meant that attempts to date HIV phylogenetically ended up in chaos, as a "dog's dinner" as he used to put it. But it was all just a front. In the end, Worobey turned out just as I had always feared he would. Shortly before Bill's partner, the Italian science writer Luisa Bozzi, died in 2004, she told me that Worobey had been opposed to me and to the OPV theory "from the start". Apparently he had told her quite early on that he was "not going to ruin [his] career for the OPV theory".
Worobey finally revealed himself in his true colours when he went back to the DRC in 2003 and managed to obtain what the two previous expeditions had not: an actual sample of viable SIV, from a chimp living in the Parisi Forest, some 110 kilometres from Stanleyville/Kisangani. This was great work, but what he did next was emphatically not. In the brief communication that he and Beatrice Hahn and Paul Sharp wrote for Nature, it was claimed that the chimps used in the OPV experiments had all come from "the vicinity of Stanleyville”, and so (they claimed) the fact that this one SIV sequence was slightly more genetically distant from HIV-1 than were the existing SIV sequences from Ptt chimps, "refuted" the OPV theory. He went on to assert that this finding "should finally lay the OPV/AIDS theory to rest." As I immediately wrote to Nature, this claim was false, not least because the Lindi chimps had been captured from a huge area of rain forest and savannah encompassing the entire northern section of the Belgian Congo, and because Lindi had housed at least one chimp that appeared likely to have been a Ptt. (At that stage I had not yet discovered the documentary evidence that at least one of the Lindi chimps had been a Ptt.) Nature ignored my letter (as well, I believe, as others submitted on the subject), and ever since then Nature and Science (and Wikipedia!) have celebrated Worobey's erroneous paper as the alleged disproof of the OPV theory.
In Tinderbox, Worobey gives a bit of background to this supposedly crucial study. On page 26 the authors state that "Hamilton and Worobey figured that the chimps who had lived at the Kisangani lab were captured nearby". (This is both clumsy and inaccurate, but what Timberg is trying to say is that Hamilton and Worobey calculated that the chimps housed at Lindi camp and allegedly used to prepare the vaccines in the Stanleyville lab must all have come from nearby areas.) The only possible source for this claim is Worobey. However, it is patently untrue.
Bill Hamilton was well aware from frequent discussions with me including those in Kisangani, formerly Stanleyville, in July 1999) that there was considerable evidence showing that the chimps had come from a huge section of the Belgian Congo that lay to the north and east of the Congo River. Some of this evidence was in the public domain; some was not. But I had shown Bill the specific places I knew about where Lindi chimps had been captured; I had actually marked these places on a 1986 edition of the Michelin map of southern Africa that I used to bring to our meetings. Moreover this fact was made clear in The River which (according to Timberg) Worobey had read, as indeed had Bill (at least twice, both in proof and its final version). On pages 581-2 and 716 I had reported that the Lindi chimps had come from several different regions spread across a 120,000 square mile swathe of northern Belgian Congo. Although I had not specified the exact capture places in the text (in a perhaps misguided attempt to safeguard chimps living in those areas), I had given full citations for the articles supporting these claims, which were available in medical libraries.
So for Worobey to claim (as this book makes clear that he has done) that the error that lies at the core of his 2004 Nature article was as much Bill Hamilton's fault as his own is not only an attempt to shift responsibility for a piece of rotten science on his part. It is also a slander against a true scientist, now no longer able to defend himself.
Perhaps the worst bit of Tinderbox features in two paragraphs right at the beginning (on pages 2-3). It includes the following passage. "We now know where the epidemic began: a small patch of remote southeastern Cameroon. We know when: within a couple decades either side of 1900. We have a good idea of how: someone caught an infected chimpanzee for food, allowing the virus to pass from the chimp's blood into the hunter's body, probably through a cut during the butchering.....It was here, in a single moment of transmission from chimp to human that a strain of virus called HIV-1 group M first appeared." So, four statements in a single page, three of which are presented as known facts and the fourth as a probable fact. But in reality, all four of these claims are hypothetical - and in my opinion (based on facts, probability and logic) each one is false. (See "Quick Guide".) Nice start, gentlemen!
After the rather sycophantic chapter about Worobey's experiences on safari with Hamilton, the narrative does improve somewhat, but because the authors are basically fleshing out a story sketched out by the bushmeat activists, some bits read much less convincingly than others. It is when in Chapter 7 the story arrives in Leopoldville at Independence in 1960, that Worobey's views once again enter the fray. HIV-1 is said to have travelled 600-odd miles from Cameroon to Leo by river-boat, and once in Leo, it has apparently spent decades spreading, and differentiating into different strains that would very soon become established as the different subtypes of the pandemic virus. Worobey is reported to have calculated that "maybe a few thousand" persons were infected with HIV by 1960. The authors state that: "Given that the symptoms of AIDS, including wasting, fevers and diarrhea, can be caused by many other tropical diseases, it's hardly surprising that nobody appeared to have noticed a new epidemic emerging in their midst." The narrative goes on to inform us that "the Belgians built - using Congolese muscle and sweat - one of the most modern and extensive transport systems in Africa." But HIV could not escape from the capital, it is claimed, until colonial travel restrictions for Africans disappeared on June 30th, 1960.
This account is breathtakingly naive, and could only have been written by people without very much knowledge of Africa, or of colonial medicine. At this point let me swiftly admit that of course I am not a doctor either. However, I have studied this subject for 25 years, and have interviewed as many witnesses to these events as I could find. I started doing this 20 years before Timberg arrived in Africa, when (to be fair) many more of these witnesses were still alive. Altogether I have interviewed more than 25 colonial doctors, including many of those who worked in internal medicine in Leopoldville in the years before and after Independence. The best-known of these, Jean Sonnet and Jean-Louis Michaux, were meticulously careful doctors who wrote very full case notes. And they (and their successors when they left around 1968) retrospectively recognised that whereas collections of AIDS-like diseases in individual patients were indeed identified after Independence in one or two scattered cases in the 60s, and a growing number in the 1970s, they were not seen before Independence in 1960. Other doctors from Leo (and indeed from other parts of Africa) agree. These were medical people who knew the syndrome well, because they were among the first to encounter it (once it had been named as AIDS) in African patients in Belgian, French and British hospitals in the early 1980s. They all agreed that AIDS simply wasn't seen before 1960.
Worobey's estimate of a "few thousand HIV-infected" in Leo by Independence isn't anything approaching a realistic figure, for if there had been that number in 1960 there would also have been a few hundred AIDS cases during the last years of the colonial era. That in turn means that AIDS (a) would surely have been recognised as an unusual new disease syndrome in the 1950s by physicians such as Jean Sonnet, and (b) would almost certainly have already begun its epidemic spread outside Leo. To suggest that HIV and AIDS for some reason holed up in Kinshasa for 30 to 70 years, spread and diversified, and then suddenly escaped after Independence, allowing the ten different subtypes to become established in different venues, is not only extremely far-fetched. It is absurd. Worobey's estimates are driven not by sound epidemiological arguments, but by the necessity of trying to propose a history to support his preconceptions of how the epidemic must have happened.
I would be surprised if in reality there were more than 100 HIV infections in Leo by 1960. And in my opinion these 100 or fewer infections existed because that number of OPV vaccinees (both Leopoldville vaccinees, and vaccinees who had migrated to Leo from other vaccinated towns) had become infected in the three previous years with differing strains of SIVs that had been recombining in the caged chimps of Lindi Camp, and in the chimpanzee tissue cultures used to prepare the locally-made batches of OPV. It is these differing SIV strains that were the true source of the various pandemic HIV-1 subtypes recognised today. It is this that explains the sudden characteristic "sun-burst" of pandemic HIV-1 subtypes that the geneticists recognise must have happened in roughly 1960.
The insistence by geneticists that they are the only ones who can properly interpret the epidemic leads straight to a Catch 22. The geneticists who write about the origin of AIDS rely on the model of the molecular clock, and base their analysis on the principle that there must have been a single index case for each epidemic. They believe that there have been four separate primate-to-human transfers to produce the four different outbreaks of HIV-1 (M, N, O and P). But the idea of a "synchronised event" (to use the phrase of Gerry Myers), a multiple transfer of, say 10 different viral strains from SIV-infected polio vaccines to a total of perhaps 50 unfortunate vaccinees, is automatically excluded by the molecular clock model.
The HIV geneticist refer led to several times above has acknowledged the truth of the claim (first made by Mikkel Schierup and Raoul Forsberg in their paper presented at the meeting on “Origin of HIV” at the Lincei Academy in Rome in 2001) that recombination affects the dating of the most recent common ancestor (MRCA), and that ignoring recombination leads one to think that the MRCA lies further back in time than it actually does. I want to state formally that I respect the integrity that this man has shown during our email dialogue, and the fact that he has been willing to accept several specific points that I have put to him (such as the one outlined above) that impact negatively on the phylogenetic/bushmeat argument. However, he still maintains that geneticists could identify such recombination, had it occurred, either in the phylogenetic trees or in the similarity plots (Simplots) that compare different viruses. But as Schierup and Forsberg argue, recombination that occurs early in the history of a virus cannot be identified retrospectively. This is especially true when one considers that the OPV theory proposes that recombination (which may have involved the SIVs of both Pts and Ptt chimps) would have first occurred in vivo, due to the co-caging and group caging at Lindi camp, then again in vitro as chimpanzee kidneys and sera were used to make tissue cultures for the vaccine batches prepared at the Stanleyville lab, and all this before the viruses in different vaccine batches were passed to human vaccinees in more than 30 different trials staged in geographically isolated sites across central Africa. This is what Gerry Myers referred to as “a synchronised event”.
It would be heartening if some of the more outspoken geneticists had the courage to admit that they might not possess the right scientific tools to analyse other possibilities regarding the early days of this epidemic. But what chance is there of that? Does anyone seriously believe that people like Hahn, Sharp and Worobey, who have invested perhaps 10 to 15 years of research time (and an awful lot of other people's money) in their theory are likely at this stage to admit that they might be wrong, and that their research methods might not be up to the task? No, these people are too far in to withdraw. Better to argue even past the point of reasonable argument that their unnaturally forced and stressed (or is it over-relaxed?) version of events has to be correct.
As for Tinderbox, there is little else that needs to be said about it. The title comes from what sounds very much like a carefully-planted quote from Worobey, in which he compares a tinderbox to wet moss: one promotes the spread of fire (or in this case viruses), the other retards it. The opening eight chapters of the book are all rather like this. They appear to have been hurriedly written and not very well researched. They read rather like a child's adventure story, and demonstrate a striking lack of investigative intent. This is a shame, because parts of the rest of the book are actually rather good.
On page 9, Timberg celebrates the fact that he (who arrived in Johannesburg as bureau chief of The Washington Post in 2004) and Halperin (a former San Francisco cab driver turned USAID epidemiologist) are "relative latecomers" to the study of AIDS. In an author's note at the front, Timberg writes that what has bonded the two men together has been " a fascination with human experience [and] a refusal to accept easy answers". Not everyone would agree with that analysis, especially the latter phrase. In my opinion, what breathes from the pages of their text is that however long these two might have lived in Africa, the views of both men are those of Americans who have been living an essentially westernised existence there. There is sympathy and goodwill by the bucket-load, but no real empathy or understanding - either for Africans or for AIDS.
Tinderbox is clearly little more than a puff (and a powder puff at that) for the bushmeat fantasists. "A naked emperor?" one can imagine Timberg asking. "Actually I never saw one. Did you see one, Daniel?"
April 25th, 2012