The Dynamics of Closure: Whither the Controversy? (1992–1995)
The Definitive Study?
Overshadowed by the news about "AIDS without HIV" were other research findings that augured less favorably for the future success of the HIV dissenters, particularly Duesberg. First, over the summer of 1992, two groups of researchers reported on the latest advances in isolating HIV from T cells using PCR technology. Instead of finding the virus in one out of every ten thousand or one hundred thousand cells, scientists were now succeeding in finding the virus in one out of every forty to two hundred cells. The following January, Dr. Ashley Haase of the University of Minnesota and other researchers reported on their refinement of PCR called "PCR in situ," which allowed them to discover HIV in 10 to 30 percent of T cells and in high percentages of other cells as well. These findings, the authors explained, were "consistent with the emerging view that HIV infection per se could contribute substantially to depletion of immune cells in AIDS."
Further relevant data were published in March, when two groups
of researchers, one led by Haase and the other by Anthony Fauci, described finding "massive covert infection" of lymphoid cells in the spleen, tonsils, adenoids, and lymph nodes during the early stages of HIV infection—ten times more virus than could be detected in blood samples at that stage of illness. "These developments," wrote Nature editor John Maddox in the Times of London, "convincingly give the lie to Duesberg's only cogent criticism of the conventional view that it is difficult to recover virus from the helper T-cells circulating in the blood." Maddox continued: "Now it seems probable that these vital components of the immune system are damaged and perhaps killed off in the relative obscurity of the lymph nodes and the other organs of the immune system."
Another piece of news (which received little notice beyond a brief report in a New York Times article devoted mainly to other topics) arrived in a presentation at the International AIDS Conference that was designed specifically to prove Duesberg wrong. The report was by Kevin Craib and other researchers with the Vancouver Lymphadenopathy-AIDS Study, one of the principal cohort studies that had been tracking HIV-negative and HIV-positive gay men for nearly a decade. The study's authors noted Duesberg's argument that no controlled study had been conducted that could truly distinguish between lifestyle risk factors and HIV infection as possible causes of AIDS. "The purpose of this analysis," wrote Craib and his associates, "was to conduct just such a controlled analysis within a cohort of homosexual men."
Looking back over their accumulated data, Craib and his coauthors investigated the incidence of AIDS-defining illnesses and changes in T-cell counts over time and tried to relate those clinical developments to HIV status, to exposure to psychoactive drugs, and to sexual behavior. They found that half of the 350 HIV-negative men in the study reported using psychoactive drugs. And about a quarter of them reported having been the receptive partner in anal intercourse with "casual partners." But despite these high incidences of (hypothesized) risk behavior, there were no signs of immune dysfunction in the seronegative group—no AIDS-defining illnesses and no drop in T-cell counts. Of 134 AIDS-defining illnesses in the full cohort, "every single one occurred in men with pre-existing evidence of HIV infection." Cofactors might determine which HIV positives contract which (if any) opportunistic infections, the authors concluded; "but claims that AIDS is caused by other exposures and not by HIV are clearly not borne out by these data."
The Berkeley epidemiologist Warren Winkelstein had already shown Duesberg unpublished figures, derived from the San Francisco Men's Health Study, that supported similar conclusions. Duesberg, however, insisted that unpublished data simply didn't count in the world of science, and he refused to acknowledge Winkelstein's findings unless they appeared in a peer-reviewed journal. Winkelstein maintained, in response, that these were really nonfindings and that scientists weren't generally in the business of publishing nonassociations. But in the end Winkelstein was induced to answer to his nemesis; and in March 1993, an immunologist named Michael Ascher, along with Winkelstein and other colleagues, published a commentary in Nature entitled "Does Drug Use Cause AIDS?" The article appeared the same week as the formal publication of Craib's study in Lancet —a double whammy that attracted the media attention that Craib's conference presentation had failed to elicit.
Ascher and his colleagues assembled a point-by-point argument that seemed intent on addressing every objection that Duesberg had ever voiced or might conceivably muster in the future. It was a strong piece of science and very much a political document, one which acknowledged as its motivation "the wide publicity attracted by [Duesberg's] assertion." Ascher and his colleagues reviewed the San Francisco Men's Health Study data for 1,027 study subjects over ninety-six months. According to interviews performed at the beginning of the study back in 1984, the 812 gay or bisexual men and the 215 straight men had used cocaine, marijuana, and amphetamines in roughly the same percentages during the two years prior to recruitment for the study. (For example, 36 percent of homosexuals/bisexuals and 39 percent of heterosexuals had reported using marijuana once a week or more.) Since 26 percent of the gay or bisexual men had been diagnosed with AIDS indicator diseases over the course of the study, one would therefore expect that roughly the same percentage of the heterosexuals would also have AIDS diseases, if drug use were indeed the cause. But in fact, there were no cases of AIDS indicator diseases in the heterosexual group.
Noting that "the clinical case definition of AIDS has been criticized as having subjective features and low specificity," the authors also presented data examining the relationship between T-cell counts over time, drug use, and one's status as antibody positive or negative. T-cell counts, after all, were "the primary pathognomonic feature of AIDS"; they were the best indicator of the health or impairment of the
immune system in people at risk of AIDS. The results were striking: The average numbers of T cells (per cubic millimeter) for the seropositives declined steadily from about seven hundred in 1984 to about four hundred in 1992, regardless of extent of drug use. But the average numbers for the seronegatives stayed the same, at one thousand to twelve hundred T cells—although ironically, the heavy and moderate drug-using seronegatives actually had slightly higher T-cell counts than the seronegative abstainers. This was a surprising finding, but it hardly helped Duesberg's case. The researchers concluded with their own challenge for their adversary: "The energies of Duesberg and his followers could better be applied to unraveling the enigmatic mechanism of the HIV pathogenesis of AIDS."
The Terms of the Debate
One of the dissenters' strongest arguments, which made an appearance in practically every published statement that any of them had made over the course of more than eight years, was that closure had been reached prematurely in 1984. A "probable cause" had been identified that was soon taken to be "the cause," yet no one bothered to prove it definitively. Instead, the world simply acted on what it took to be the truth. By the mid-1990s, however—particularly as a result of data from long-running cohort studies—the state of the evidence looked rather different. Indeed, one might even argue that the dissenters' challenge and the ensuing controversy had served to promote clearer argumentation on behalf of the HIV hypothesis. In that sense, ironically, the dissenters may have helped to reinforce the dominance of the position they opposed.
Does this mean that the end of the controversy was in sight? That would have been an overly hasty conclusion, and one that ignored recognized pitfalls along the path to closure in scientific controversies. In their consideration of the question of how scientific controversies end, H. Tristram Engelhardt and Arthur L. Caplan have noted the danger in assuming that complex disputes can be adjudicated through straightforward, rational means. The notion that a controversy can be "settled" presupposes prior agreement on "(1) how to acquire evidence relevant to the dispute and (2) how to reason with the evidence in order to reach a rationally defensible conclusion that will resolve the controversy." In fact, differing perceptions of what "the controversy" is and what sort of evidence might "settle" it are themselves often stakes in a controversy. This is all the more likely to be true,
Engelhardt and Caplan have observed, when "stakeholders in the debate belong to … different communities with different appreciations of the evidence at stake" or to "competing social groups" with opposing political and ethical agendas.
How does one prove disease causation? Without agreement on this basic question, there was little chance of consensus about the relation of HIV to AIDS. In the early 1980s, many researchers, including even Gallo, seemed to have Koch's postulates at least generally in mind as they sought to isolate the virus, to induce the syndrome in primates, and to trace transmission of AIDS through infected blood products (see chapter 2). But when challenged by Duesberg, researchers argued that Koch's postulates were rarely satisfied in practice, especially for viruses; and they pointed to well-known examples.
As Duesberg pressed his case, both sides moved toward the position that specific epidemiological evidence—especially controlled longitudinal data—is what could settle the question. But where epidemiologists such as Winkelstein insisted that the existing data from cohort studies already provided a definitive answer, Duesberg maintained that a meaningful cohort study had not yet been performed: The existing studies were created with the goal of studying the progression of "HIV disease." They were not set up to determine what causes AIDS, let alone to study drugs as a possible cause, and, Duesberg claimed, the data that had been collected about participants' drug use were superficial and inadequate. Neither Winkelstein nor the Vancouver group asked about drug use occurring as far back as ten years prior to entry into the study; nor had any of the cohort studies included blood tests for drugs, which Duesberg claimed were necessary to verify if the study participants were answering truthfully. From the perspective of Duesberg and his allies, these aspects of data collection rendered the cohort studies worthless in adjudicating between causal hypotheses. Duesberg's deconstruction of the cohort studies demonstrated clearly what science studies scholars have repeatedly argued: in situations of controversy, experiments alone cannot be expected to "settle" the dispute. "The problem with experiments," according to Harry Collins and Trevor Pinch, "is that they tell you nothing unless they are competently done, but in controversial science no-one can agree on a criterion of competence." Moreover, any experiment is—admittedly—an artificial stand-in for real-world conditions. Does the experimental situation adequately represent and capture reality? Nothing inherent in the experiment itself forces the observer to accept it as such.
Both sides remained convinced of the validity of their claims; each
side saw the other as dogmatically defending its position by means of a constant renegotiation of the rules. From Duesberg's standpoint, the defenders of orthodoxy were always "moving the goalposts" whenever its predictions were proved false; from the vantage point of the dominant position, the dissenters were forever cooking up newer and stricter criteria of proof that their opponents were then expected to meet. "Like a child who questions every answer with another 'why,' he plays a game which will never end until he says it is over," was how Martin Delaney characterized Duesberg's methodology in late 1992. But perhaps the real mistake was to imagine that there is one "scientific method" or set of "rules of evidence" upon which everyone could unproblematically agree. Indeed, if the controversy has demonstrated anything, it is that scientific "rules of the road" like Koch's postulates or epidemiological cohort data, far from being unchallengeable benchmarks, may be as subject to interpretation and debate as the empirical phenomena they are invoked to explain.
Most likely there is no scientific test that would settle the causation controversy to the satisfaction of all sides. A different sort of proof probably would spell the end of the controversy, however: if the AIDS establishment were to succeed in finding an impressively successful antiviral drug or vaccine, it is unlikely that anyone would continue to pay much attention to the dissidents. They would simply fade from the scene. This, however, is not entirely logical, for, as René Dubos has argued, it can be dangerous to infer medical etiology "backwards" from treatment effectiveness: "While drenching with water may help in putting out a blaze, few are the cases in which fire has its origin in a lack of water." Yet at its root the support for the dissenters stems precisely from disappointment with the establishment's lack of therapeutic success, as the dissenters themselves have been acutely aware. Michael Callen told a Canadian television program in 1989: "It seems to me that if … you've got a chemical that is anti-retroviral … and you give it to people who have a disease that you claim is caused by a retrovirus and they don't get better, that would tend to suggest—at a minimum—that something a lot more complicated than having or not having HIV is going on." The more time that has passed without a cure being discovered, the more persuasive has been the argument that the AIDS establishment has "produced nothing" and—by extension—their etiological hypothesis must be bankrupt. From a logical standpoint, this, of course, follows even less obviously than the converse claim that the existence of a successful treatment would verify the
causal hypothesis. But the argument has a rhetorical power that cannot be denied. "As long as there's no cure, this will lurk," Robert Gallo reflected in 1994.
Meanwhile, the struggle has gone on, and the HIV dissenters have continued to make the news. By now several of the principals have told their stories in book-length detail. Root-Bernstein's Rethinking AIDS: The Tragic Cost of Premature Consensus , published by Free Press, weighed in at 512 pages and included 100 pages of notes. It reviewed the anomalies in the HIV hypothesis, discussed a range of multifactorial theories of AIDS, and delved deeply into theories of autoimmune mechanisms. Lauritsen's The AIDS War: Propaganda, Profiteering and Genocide from the Medical-Industrial Complex , a reprinting of many of his articles, likewise ran nearly 500 pages. Ellison and Duesberg encountered more obstacles with their book, which was accepted but then dropped by two different publishers. By 1994 the book fell victim to a bitter feud that erupted between the two authors and led to a parting of the ways. Working alone, Duesberg finally published the book Inventing the AIDS Virus in 1996. In addition to 463 pages of new text, the book included reprints of several of Duesberg's scientific articles; it also featured a foreword by Kary Mullis. In this extensive review for the general reader, Duesberg summarized the views he had been putting forward for nearly a decade—not only on the HIV hypothesis and the drug-AIDS hypothesis, but on the status of the field of virology and the politics of science in general.
The ranks of the dissenters were depleted in December 1993 by the death of Michael Callen, who had lived with AIDS for nearly twelve years and played a central role in a range of activist projects. Other dissenters have continued to engage public attention. Notable occasions have included the appearances by Duesberg, Kary Mullis, and Philip Johnson at a special panel of the annual meeting of the Pacific Division of the American Academy for the Advancement of Science; a Florida physician's public self-injection with apparently HIV-infected blood to draw attention to "the greatest scam ever perpetrated"; and a freshman Republican Congressman's attempt to force government scientists at the NIH and CDC to consider the view that HIV does not cause AIDS.
Perhaps most noteworthy, however, was a high-profile "Special News Report" called "The Duesberg Phenomenon," published in Science in late 1994 and taking up an extraordinary eight pages of the
premier journal. "Because the Duesberg phenomenon has not gone away and may be growing," wrote Jon Cohen, the AIDS reporter for Science and the author of the news report, "Science has decided this was a good time to examine Duesberg's main claims." Cohen's strategy was to tackle Duesberg's arguments head-on; in a series of sidebars, he considered each of the critical test cases. Although scrupulous in presenting Duesberg's views, the article suggested that Duesberg was wrong on all counts: being HIV positive was the key variable associated with HIV-related illness for hemophiliacs; Koch's third postulate had been satisfied by the accidental infection of three laboratory workers, all of whom developed symptoms of immune suppression prior to taking any antiviral drugs; and large, recent studies of extended AZT use showed no evidence that the drug caused AIDS. Not surprisingly, this intervention ignited yet another wave of publicity: though Cohen may have intended exactly otherwise, his article propelled the Duesberg phenomenon onward. Indeed, a long article in The Scientist published in March 1995, which led off with the special report in Science , was entitled "A Controversy That Will Not Die: The Role of HIV in Causing AIDS."
Years after the publication of his article in Cancer Research , Peter Duesberg has settled in for the long haul. The HIV controversy has become his life; his office on the Berkeley campus is the war room of a campaign waged simultaneously on multiple fronts. Duesberg maintains a frenetic pace, juggling reporters and visitors between poring over the latest AIDS publications and preparing his latest rebuttals. Though he might not agree with much that Anthony Fauci has to say, he would doubtless concur with one prediction that the NIH scientist made back in 1989: "I am probably going to be answering Peter Duesberg for the rest of my life."