The Plan of the Book
Two sets of controversies demonstrate with particular force the centrality of credibility struggles in the constitution of scientific knowledge about AIDS: debates about the causes of the syndrome and debates about treatments . The analyses of these two controversies occupy the first and second halves of this book, respectively. These debates strike at some of the central questions that confront biomedical science: What is AIDS and what causes it? How can its effects be curtailed?
Part One: The Politics of Causation
Who could doubt that HIV, the human immunodeficiency virus, causes AIDS? That proposition has been the accepted scientific wisdom since the mid-1980s, after several groups of researchers reported finding a previously unknown virus in the blood of AIDS patients. It is a conclusion endorsed by preeminent virologists, immunologists, epidemiologists, and clinicians; by the World Health Organization and the Centers for Disease Control and Prevention (CDC); and by prominent AIDS service and advocacy organizations. In the mainstream media, HIV is casually referred to as the "AIDS virus"; among insiders, AIDS is increasingly understood as simply the end stage of "HIV disease." The claim that HIV is the etiological agent in the Acquired Immunodeficiency Syndrome is the guiding assumption behind billion-dollar programs for HIV antibody testing, antiviral drug development and treatment, and vaccine research around the world. It is the cornerstone of "what science knows about AIDS."
Yet the search for the cause of AIDS took many twists and turns before settling on HIV. Indeed, the notion that AIDS might be caused by a previously unknown virus was initially a relatively unpopular one. Beginning from a zero point of near-total uncertainty, competing groups of scientific claims-makers, under the watchful gaze of interested segments of the public who sought to establish "ownership" over the epidemic, advanced various hypotheses. Then, between 1984 and 1986, a bandwagon formed behind the proposal that a particular virus, eventually named HIV, was the causal agent.
Nevertheless, in the late 1980s and early 1990s, some years after the discovery of HIV and the large-scale implementation of social policy based on it, the markers of controversy abounded. Symptomatic were debates about "cofactors" needed to cause disease, investigations into the mysteries of "pathogenesis" (Just how does HIV cause disease?), and scares about cases of an "AIDS-like" illness in HIV-negative people. Most astounding of all have been the claims of Peter Duesberg, a molecular biologist at the University of California at Berkeley and a member of the elite National Academy of Sciences. Beginning in 1987 in an article in Cancer Research and subsequently in articles in publications such as Science and the Proceedings of the National Academy of Sciences , Duesberg has maintained that HIV is a harmless passenger in the AIDS epidemic, "just the most common among the occupational viral infections of AIDS patients and those at risk for AIDS, rather than the cause of AIDS." He argues that there is no solid evidence establishing a causal role for the virus and, furthermore, that a retrovirus such as HIV simply cannot cause a syndrome like AIDS. Instead, Duesberg's current alternative hypothesis is that "the American AIDS epidemic is a subset of the drug epidemic," attributable primarily to long-term consumption of recreational drugs and secondarily to what Duesberg calls "AIDS by prescription"—the toxic effects of the medication azidothymidine (AZT), widely prescribed to fight HIV infection.
Duesberg is only one of a number of researchers, doctors, and activists who have cast doubt on the "HIV hypothesis," but he has attracted by far the most attention. Duesberg's claims have prompted dozens of articles and communications in scientific journals and several hundred articles and letters in the mainstream English-language press. In 1994, Science , one of the most important general science journals in the world, devoted eight pages to the "Duesberg phenomenon." The story has found its way into Naturwissenschaften and the Gaceta Médica de México; the Los Angeles Times , the New York Times , and the Times of London; National Public Radio and Penthouse;[ 132] the position papers of an AIDS advocacy organization and the columns of a pop music magazine; and perhaps every gay and lesbian news source in the United States. Reporters are quick to stress Duesberg's impressive credentials. He is frequently cast as a "heretic" who, like Galileo, has been excommunicated by dogmatic proponents of "orthodoxy" less interested in truth than in their hold on the faithful. And reporters are often quick to mention that Duesberg has declared himself in principle "quite happy to [be]
publicly injected with HIV." Many scientists who think Duesberg is dead wrong are made apoplectic by the mention of his name. "I'm so tired of hearing the Peter Duesberg crap about HIV," said Donald Francis, a prominent AIDS researcher formerly with the CDC, to an audience of one thousand at a 1992 public forum on AIDS in San Francisco. "News reporters looking for an AIDS angle should look for another story…. The disease is caused by the virus, dammit, and the press should understand that."
To describe the construction of facts such as "HIV causes AIDS," sociologists of scientific knowledge have adopted the phrase "black box." As Bruno Latour explains, the concept is borrowed from cybernetics, where black boxes are used in diagrams as a quick way of alluding to some complex process or piece of machinery: if it's not necessary to get into the details, one just draws the box and shows the input and the output. Then no one has to worry about what goes on inside the box itself, and the nonexpert may never even realize just how messy the inside really is. Scientific facts are similar: masked beneath their hard exterior is an entire social history of actions and decisions, experiments and arguments, claims and counterclaims—often enough, a disorderly history of contingency, controversy, and uncertainty.
Scientists strive to "close" black boxes: they take observations ("The radioactive isotope count that indicates the presence of reverse transcriptase, an enzyme associated with retroviruses, rises over time in specially prepared lymph tissue from a person with an illness believed to be AIDS-related"), present them as discoveries ("A novel human retrovirus has been grown in T-lymphocytes of AIDS patients"), and turn them into claims ("The probable cause of AIDS has been found") which are accepted by others ("HIV, the putative cause of AIDS,…") and may eventually become facts ("HIV, the virus that causes AIDS,…") and, finally, common knowledge , too obvious even to merit a footnote. Fact-making—the process of closing a black box—is successful when contingency is forgotten, controversy is smoothed over, and uncertainty is bracketed. Before a black box has been closed, it remains possible to glimpse human actors performing various kinds of work—examining and interpreting, inventing and guessing, persuading and debating. Once the fact-making process is complete and the relevant controversies are closed, human agency fades from view; and the farther one is from the research front, the harder it is to catch glimpses of underlying uncertainties. It then becomes difficult to ask, Was the examination accurate? Was the interpretation
defensible? Was the persuasion logical? Those who want to challenge a claim that has been accepted as fact must effectively "reopen" the black box.
What are the dynamics of fact making when science is closely scrutinized by attentive spectators? What are the processes by which black boxes are closed and reopened when scientific arguments become the stuff of news reports and street conversations? There are examples of important controversies in science—the debate over continental drift is one —that barely get any airplay in the "outside" world. AIDS is something else again. With millions of people around the world believed to be infected with HIV, the human stake in the causation controversy is gigantic, immediate, and inescapable. It should therefore come as no surprise that the cast of characters in AIDS debates is diverse. A full-fledged inquiry into the controversy immediately bursts us out of the "scientific field" narrowly construed. It forces an examination of the ensemble of social actors, with varying and conflicting social interests, who at different points have struggled to assert credible knowledge about the epidemic or to assert their capability to weigh and evaluate such knowledge.
At a different level, the causation controversy reflects a struggle for "ownership of" and "democracy within" science. An agenda has emerged, well expressed in the words of writer Jad Adams, one of the "HIV dissidents" and author of AIDS: The HIV Myth: "Ultimately, expert advice must be evaluated by the people who are not experts—politicians, journalists, and the public. This is part of democratic life and a scientist has no more right to exclusion from public scrutiny than a treasury official." In the intervention of laypeople in debates about the causes of AIDS, claims about causes are interwoven with claims about the very right to intervene. It makes sense that the opponents of the orthodox position on causation so frequently take aim at what they call the premature "rush to judgment" in 1984 on the question of causation: from their perspective, this moment represented the stifling of democratic openness of opinion and the authoritarian imposition of closure. In many ways the debate has become a debate about closure —that is, a debate about when and how scientific controversies end. But concerns about closure in this case break down into a number of important dimensions: epistemological (When is causation proven?), methodological (How should rival theories be weighed and compared?), empirical (Was closure arrived at too early? What conclusions did the evidence permit in 1984 and what conclusions are
reasonable today?), and most notably, political (Who decides? Which social actors are qualified or entitled to participate in the process of establishing the scientific knowledge about AIDS?). In other words, the controversies about what causes AIDS are simultaneously controversies about scientific controversies and how they should be adjudicated—controversies about power and responsibility, about expertise and the right to speak. As frames of knowledge and belief about AIDS have become fixed in place, a range of social actors have engaged in credibility struggles to defend, refine, subvert, overturn, or reconstruct those frames.
Conceived as a series of multilateral credibility struggles, the controversy surrounding the causation of AIDS raises a number of important questions that I consider in the chapters that make up part one:
How did the hypothesis that was initially considered relatively unlikely—that AIDS is caused by a previously unknown virus—come to supplant more popular alternatives?
What were the processes by which the HIV hypothesis, once formulated, became "black-boxed" and achieved the status of fact—among doctors and scientists, in the mass media, in gay communities, and in the AIDS movement? To what extent was there dissent, and how was it manifested?
How is it that Duesberg, initially presenting views that had minimal credibility among established AIDS researchers and mainstream AIDS organizations, has been able to attract allies and adherents not just among sectors of the AIDS movement but, eventually, within legitimate scientific circles as well?
How has Duesberg been able to accomplish this when others with somewhat similar views have been marginalized?
My reconstruction of the history of claims-making necessarily ranges widely. It is not just that the controversy extends across a range of scientific disciplines—virology, epidemiology, immunology, molecular biology, pharmacology, toxicology, and clinical medicine. The controversy also spins off into a series of more general debates about the nature of disease and the methods of scientific reasoning: Do diseases typically have a single cause or multiple causes? Are established rules of scientific proof inviolable, or are they subject to revision as scientific knowledge changes and technologies improve? When anomalies are found that appear to falsify an existing hypothesis, when should the
hypothesis be scrapped and when is it proper scientific procedure to work with the hypothesis, tinkering with it so that it can account for the anomalies? Does normal, peer-reviewed "establishment science" produce the best results in the end, or do the truly revolutionary findings come from the mavericks and iconoclasts who challenge, or work outside of, the system?
And finally, as the controversy has expanded from one social arena to the next, it has never been articulated in a vacuum, separate from other social concerns. On the contrary, a variety of apparently tangential beliefs and values have spilled over into (indeed, partially constitute) the AIDS etiology debates. These beliefs and values include divergent attitudes toward homosexuality, promiscuity, and drug use; inferences that link the causes or origins of a disease with theories of social blame; and assumptions about whether illnesses are attributable primarily to microbes, lifestyles, or societies. In short, the AIDS causation controversy is inexplicable outside of the larger context of how AIDS has been constructed as a social problem against the backdrop of contested attitudes about scientific medicine.
Part Two: The Politics of Treatment
Find the cause, then find the cure: this is the mission of biomedicine in a nutshell. But how does it work in practice? What are the social processes that bring a therapy from laboratory bench to medicine cabinet? Who decides what treatment strategies to pursue or how to develop and test medications? Exactly what does it mean to say that a treatment "works"? Like debates about the causes of AIDS, claims and counterclaims about treatments involve fervent struggles for credibility—struggles waged in the shadow of towering uncertainty and driven by urgent need. Progress, and power, derive from the ability to submit credible answers—to push back the bounds of uncertainty, to offer something that "helps," to voice what is "known."
The actors in this drama are as varied as the interests that motivate them and the values that animate them—the researchers hoping to hit on breakthroughs in the basic or applied sciences of AIDS research; the pharmaceutical and biotechnology companies whose stock values might fluctuate by millions of dollars, depending on the latest reports about the successes or failures of their products; the medical professionals who must translate inconclusive and contradictory research findings into workable, day-to-day clinical judgments; the regulatory agencies and advisory bodies that serve as "gatekeepers," ruling on
the safety and efficacy of new therapies; the patients who consume the drugs and populate the clinical trials; the reporters and journalists who interpret scientific research findings to various segments of the public; and, of course, the activists who police the whole process and offer their own interpretations of the methods and the outcomes. In the late 1980s, "treatment activism" emerged as the forward wedge of the multifocal AIDS activist movement in the United States, widely hailed—and sometimes damned—for its ingenuity, brashness, aptitude, and muscle.
"There's no doubt that they've had an enormous effect," commented Dr. Stephen Joseph in 1990, soon after leaving the post of New York City health commissioner. "We've basically changed the way we make drugs available in the last year." While the activist impact on the regulatory procedures of the Food and Drug Administration (FDA) has been widely publicized, this remains just one of many items on treatment activists' agenda for the reformation of biomedical science. As the National Research Council of the National Academy of Sciences put it (in a 1993 report otherwise noteworthy for its skepticism about the transformative effects of the AIDS epidemic on U.S. society): "Every aspect of the process by which new pharmacologic agents [are] identified, evaluated, regulated, and allocated [has been] tested by the exigencies of [this] epidemic disease. Questions basic to the epistemologic foundations of biomedicine—questions of verifiability, reproducibility, proof, variability, safety, and efficacy—[have all been] subject to debate and reevaluation."
Treatment activists have been pivotal in this rethinking of biomedical truth-making. They have challenged the formal procedures by which clinical drug trials are designed, conducted, and interpreted; confronted the vested interests of the pharmaceutical companies and the research establishment; demanded rapid access to scientific data; insisted on their right to assign priorities in AIDS research; and even organized research on their own, with the cooperation of allied professionals. Starting out on the margins of the system, treatment activists have pushed their way inside, taking their seats at the table of power. Activists now sit as full voting members of the NIH committees that oversee AIDS drug development, as invited participants at the FDA advisory committee meetings where drugs are considered for approval, as members of federal review panels that consider proposals for research grants, and, at local levels, as representatives on the review boards that approve clinical research at hospitals and academic centers.
In addition to moving inward, they have pursued an evolution "backward," as treatment activists themselves have noted. Beginning with a focus on the end stage of the drug development process, they have worked their way back toward earlier and earlier moments—"from drug approval at the regulatory level of [the FDA], to expanded access for drugs still under study …, to the design and conduct of the controlled clinical trials themselves. …" Most recently, several prominent treatment groups have pushed back even further, to promote, monitor, and criticize the directions of basic AIDS research—the "pure science" investigations in immunology, virology, and molecular biology that are considered the necessary prelude to the applied work of developing and testing specific therapies.
The vigorous participation of self-educated activists—and more broadly, the rise of knowledge-empowered communities that monitor the course of biomedical research—has had momentous effects on the development of AIDS treatments. These developments have transformed the procedures by which drugs are tested, the ways in which test results are interpreted, and the processes by which those interpretations are then used in the licensing of drugs for sale.
The Conduct of Clinical Research.
In the postwar era, the assessment of therapies has been linked to the techniques of the randomized clinical trial. Such trials provide crucial "hard data" about treatment effects, but also obscure political decisions about how to measure the risks and benefits of a drug, cloaking them in the aura and mystery of objective science. Widely considered the pathway to objectivity in modern biomedical research, clinical trial results in practice can be subject to enormous amounts of interpretative flexibility. Precisely because the stakes are often high—both in human lives and in stock market values—deciphering clinical trial findings can prove not only a contentious process, but also a highly public one.
Clinical trials are also a form of experimentation that requires the consistent and persistent cooperation of tens, hundreds, or thousands of human beings—"subjects," in both senses of the word, who must ingest substances on schedule, present their bodies on a regular basis for invasive laboratory procedures, and otherwise play by the rules, known more formally as the study protocols. From the standpoint of the researcher, ensuring the cooperation of research subjects is a complicated endeavor because these "bodies" talk back: subjects participate or don't participate and comply with the study protocols or
not, depending on their own perceptions of what works and what doesn't, how desperate their own health situation is, and what options are open to them.
It has recently been argued that the history of clinical trials needs to be rewritten to, in effect, "bring the patient back in"—to demonstrate how the capacity to construct knowledge through this particular technique is both enabled and constrained by the research subjects and the resistance they present to the epistemic goals of the clinical investigators. In fact, the AIDS epidemic should be considered a decisive turning point in this revisionist history. AIDS trials are distinctive not only because of the militancy of many of the patients, but because their representatives have mobilized to develop effective social movement organizations that evaluate knowledge claims, disseminate information, and insert laypeople into the process of knowledge construction. The activist representatives of AIDS patients not only facilitate the flow of information to and among them, but also press demands about what should be studied in the first place and how the research protocols should be worded. Highly technical details such as the entry requirements for trials, the types of controls employed, and the endpoints to be used in studies have all been the subject of vociferous debate. Such developments pose substantial complications for the "politics of therapeutic evaluation."
The Interpretation of Studies.
"You can't reproduce the real world in a … clinical study," acknowledges Dr. Douglas Richman, a prominent AIDS researcher at the University of California at San Diego. "The hope is that you can define things in such a way that you can get some interpretable data in which the bias is sufficiently limited [so] that it's meaningful and it's applicable to other situations. …" As Steven Shapin expresses it, any laboratory experiment has credibility only insofar as it is taken to "stand for" some actual conditions in the "real world": for example, "when Robert Boyle put a barometer in the air-pump and then exhausted the air, its behavior was meant to stand for what would happen were one to walk a barometer up to the top of the atmosphere." But the extent to which the experiment adequately represents reality is always subject to negotiation—and open to deconstruction. The effect of activist interventions into questions of research design and interpretation has been precisely to "denaturalize" clinical trials—to call the objectivity of the methods into question, to reveal their "artifactual and conventional" status, and to make the results of given trials more open to question.
Few people, including practicing physicians and many academic researchers who conduct clinical trials, can entirely follow all the statistical arguments that constitute the formal evidence invoked in favor of, or in opposition to, a given treatment. Most players therefore become adept at reading the signposts: Where was the study published? Who conducted it? Was it peer reviewed? Is anyone criticizing it? Are there any methodological flaws or "gray areas" that have been pointed out? Has the FDA acted on it? Has the NIH issued treatment guidelines? Do I know of doctors who are prescribing this drug? How are their patients doing? The social power of a study depends considerably on these markers of credibility, and their absence can cause problems for the acceptance of the study's findings.
Into this complex field of claims and counterclaims, markers and precedents, signals and responses, enter the AIDS activists. Activist participation has done nothing less than change the ground rules for the social construction of belief —the varied processes by which different groups and institutions in society come to believe that a given treatment is "promising" or "disappointing," "effective" or "junk," "state of the art" or "passé." Activists have become proficient at interpreting the credibility of AIDS trials, and they have educated their base communities about how to scrutinize newspaper reports of "miracle cures" and journal articles about the "definitive" clinical trial. Activists in turn have promoted their own assessments: that study makes sense; this drug seems to be working. They have argued for rethinking the risk-benefit calculus for life-threatening illnesses, and they have pushed for the rights of patients to accept greater risk in deciding whether to try experimental treatments.
The Politics of Risk and Regulation.
In a world that depends heavily on specialized expertise, decisions about risks increasingly are adjudicated by impersonal organizations and institutions. But particularly in cases of controversy and politicization, it becomes harder to "contain" such decisions within normal organizational routines. Thus the regulatory hearings that consider evidence from clinical trials in order to license pharmaceutical products for sale are often heated sites for the negotiation of credibility, risk, and trust. "Regulatory science," as Sheila Jasanoff calls it in her study of agencies such as the FDA and the Environmental Protection Agency, is indeed a legitimate variety of scientific enterprise, but it is a very particular variety—and at least in the United States, a particularly adversarial one at that. Regulatory science differs from research science in its goals, its
institutional locus, its formal products, its time frame, and its accountability.
A distinctive difficulty of regulatory science is that everyone involved in making assessments speaks a somewhat different language. The statistician wants to know if the "null hypothesis" of "no treatment effect" has been disconfirmed to a sufficient degree of statistical certainty. The clinician wants to know if her patients' symptoms show improvement. The pharmaceutical manufacturer is concerned with liability and profit margins. And the regulatory official assesses "safety and efficacy" by measuring compliance with statutory and administrative requirements. Patients and their activist representatives want to know if a drug "works": at times they may demand certainty from institutions ill-equipped to provide it; at other moments they may insist on their willingness—indeed, their right —to freely assume the risks of uncertainty and ingest substances about which researchers and regulators have doubts. Should the social priority be "access" or "answers"—rapid approval of experimental therapies or careful consideration of the accumulation of evidence? Are these goals in conflict or can they be advanced simultaneously? Regulators, researchers, doctors, patients, and activists have all held different opinions on these questions—opinions that in some cases have shifted markedly over the course of the past decade.
A small number of drugs, many of them chemical cousins of AZT, have been licensed in the United States as antiviral agents effective against HIV infection or AIDS. (Other drugs have been licensed to fight opportunistic infections and neoplasms that are characteristic of AIDS—the infections and cancers that afflict people with weakened immune systems.) It is universally agreed that none of these drugs is a cure for AIDS. Beyond that, the drugs are shrouded in controversy, and AZT, the drug most widely prescribed throughout the late 1980s and early to mid-1990s, has seen its star rise and then fall. What is "known" and "believed" about these drugs, by whom, and where?
The answers to these questions demonstrate how certainty and uncertainty about treatments—much like beliefs about causation—emerge out of multilateral struggles for credibility. I analyze these struggles by tracing the first decade of AIDS antiviral drug development, from 1984 to 1995, and emphasizing the following issues:
What were the approaches to treatment that emerged upon the "black-boxing" of the HIV hypothesis? How have treatment strategies
changed in relation to advances in virology and immunology?
How did AZT become credible as the treatment of choice for AIDS and for lesser HIV infection? How did a bandwagon form behind the promotion of this drug? Why was the drug recommended for many asymptomatic (outwardly healthy) HIV-infected people in 1989, and why did its value for those same patients come into doubt by 1993? Who has supported AZT, who has questioned it, who has labeled it a "poison," and why?
How did AZT's "cousins" ddI, ddC, d4T, and 3TC become the beneficiaries of novel accelerated mechanisms of drug testing and licensing? What is known about the long-term efficacy of these drugs and about newer, perhaps more promising drugs such as the "protease inhibitors"?
What impact has the "democratization of research" had on the assessment of credibility and the construction of belief about drugs? Specifically, what have been the consequences of lay participation on NIH committees and on community advisory boards for clinical trials? How has the development of a sophisticated activist press that evaluates treatments and regulatory policies affected the research and regulatory processes?
How have activists constructed their credibility as scientific actors in this domain? How have they pressed their critiques of the methods employed in the design and interpretation of clinical trials? How has the engagement with clinical and basic science affected the organization and identity of the movement?
In the debates surrounding treatments, just as in those concerning causation, we see participants engaged in disputes about the meaning of evidence that simultaneously are disputes about the standards of evidence—about what counts as proof and, crucially, who decides. In this respect, there is an underlying commonality between part one and part two of this book. (Indeed, many of the same players surface in both). At the same time, the causation and treatment controversies are in certain respects quite different, and they reveal distinct mechanisms for the establishment of credibility in science. In both cases, the AIDS movement, broadly construed, has played an important and visible role. But in the causation controversies, the most publicized challenges to mainstream views have come from highly credentialed researchers, and representatives of the AIDS movement have enjoyed more success
in assessing the claims of others than in asserting their own. By contrast, the crucial voices of heterodoxy in the treatment controversies have been those of lay activists.
Given the nature of scientific research, there is a certain logic to this difference. The investigation of the etiology and pathogenesis of illness is closer to the realm of "pure science"—the heavily-defended "core" of scientific practice to which few outsiders can successfully gain entry. By contrast, the investigation of treatments—and in particular the establishment of treatment efficacy through the mechanism of the clinical trial—is in part an "applied" science, located more on the "periphery" of scientific practice. It is more easily accessible to members of the patient community, whose participation in the process is indeed essential and who therefore have an immediate claim to a stake in the process and a basis for the development and assertion of their own expertise. A comparison of AIDS causation controversies and treatment controversies is therefore instructive, for it demonstrates how different kinds of scientific debates generate different possibilities for the manufacture of scientific expertise and credibility.