"Amid the dozens of books about AIDS, one stands out—Impure Science. . . . Epstein has documented the fast-moving history of the epidemic's first years in an eloquent, readable narrative. . . . Intelligent and original."
Impure Science: AIDS, Activism, and the Politics of Knowledge / Edition 1 available in Paperback
"As the AIDS movement is showing, people with diseases can play a profound part in saving themselves . . . A perceptive and useful analysis of this revolution in the democratization of medicine."
"A monumental book to read and ponder."
"A study marked by scrupulous attention to detail that is at the same time almost breathtaking in its scope and probing in its analysis. It is at once a fine contemporary history of science, a sociology of knowledge, and an account of the emergence and fate of a social movement driven by rage and passion."
"Lucid, balanced, and impressively well-documented."
"For those seeking insights into what surely is the greatest medical story of our times, Impure Science provides a rich lode of contextual material from which to consider howe we got here."
Read an Excerpt
Impure ScienceAIDS, Activism, and Politics of Knowledge
By Steven Epstein
University of California PressCopyright © 1998 Steven Epstein
All right reserved.
The Nature of a New Threat
The Discovery of a "Gay Disease" (1981-1982)
When a puzzling new medical syndrome was first reported to be afflicting—and killing—young gay men in certain cities in the United States, there was no particular reason to expect that the cause might be a previously unknown virus. Nor did the deaths immediately take on any great medical significance. Michael Gottlieb, a young immunologist at the teaching hospital of the University of California at Los Angeles, began seeing such cases in late 1980 but found that he couldn't spark the interest of the New England Journal of Medicine , the most prestigious medical journal in the country, later to publish hundreds of articles on AIDS. In early 1981 the New England Journal 's editor instead referred Gottlieb to the U.S. Centers for Disease Control (CDC), the federal agency in Atlanta, Georgia, responsible for tracking diseases and controlling their spread.1
The CDC's first report, published in its Morbidity and Mortality Weekly Report in June 1981, noted only that five young men in Los Angeles, "allactive homosexuals," had been treated over the course of the past year for Pneumocystis carinii pneumonia (PCP).2 Two of the men had died. The microorganism that causes PCP is ubiquitous but is normally kept easily at bay by the body's immune system; therefore cases of PCP were exceedingly rare, restricted to people who wereimmunosuppressed because of medical treatment (such as chemotherapy) or who for other reasons had severely malfunctioning immune systems. The CDC report zeroed in on the question of sexuality—"the fact that these patients were all homosexual"—to put forward two tentative hypotheses: that the PCP outbreak was associated with "some aspect of a homosexual lifestyle" or with "disease acquired through sexual contac." However, "the patients did not know each other and had no known common contacts or knowledge of sexual partners who had had similar illnesses."
A few weeks later, the CDC reported twenty-six cases (twenty in New York City and six in California) of young homosexual men suffering from Kaposi's sarcoma, a rare form of cancer normally found in elderly men. At least four of the men also had cases of PCP; eight of them had died.3 On the basis of this report, Dr. Lawrence Altman, medical reporter for the New York Times , wrote a short article about the cases of cancer in homosexuals.4 Appearing deep inside the newspaper on page A-20, the article sounded what would become one of the most common themes in mainstream media coverage of the epidemic: "The reporting doctors said that most cases had involved homosexual men who have had multiple and frequent sexual encounters with different partners, as many as ten sexual encounters each night up to four times a week." Soon after, Dr. Lawrence Mass, health writer for the New York Native —the most widely read gay newspaper in New York City and one of only a few to have a national readership—also addressed the question of promiscuity. In an article about "Cancer in the Gay Community," Mass wrote: "At this time, many feel that sexual frequency with a multiplicity of partners—what some would call promiscuity—is the single overriding risk factor. …"5
Mass's article also explored a range of possible explanations for what he called (in quotes) "the gay cancer," including "an infectious or otherwise cancerous agent," but he noted that the "current consensus of informed opinion is that multiple factors are involved in the present outbreak of Kaposi's sarcoma among gay males." He quoted Dr. Donna Mildvan, chief of infectious diseases at Beth Israel Medical Center, who reported a colleague's belief that the outbreak of illnesses "has to do with the bombardment, the clustering of a whole range of infectious diseases among these patients which may be exhausting their immunodefensive capacities." And he cited Dr. Alvin Friedman-Kien, a professor of dermatology and microbiology at New York University Medical Center, who had examined some of the Kaposi's sarcomapatients and who had speculated about the possible role of amyl nitrite or butyl nitrite inhalants. These inhalants, street drugs that were sold legally and were popular in gay male communities at the time, were often called "poppers" because consumers would pop open the packaging to release the fumes, which were then inhaled to produce a "rush" or to intensify orgasm. Nitrites were believed to have immunosuppressive effects. On the other hand, they had been prescribed to cardiac patients for years, and no unusual cases of PCP or Kaposi's sarcoma had ever been reported in that population.
By the beginning of 1982, a series of more detailed reports in medical journals such as the New England Journal6 was available as a source of additional information and speculation for researchers and medical practitioners, and for translation into the media, particularly the gay press. Researchers agreed that the telltale marker of these cases of immunosuppression was a deficiency in the numbers of "helper T cells"—or in other accounts, an abnormal ratio of helper T cells to suppressor T cells—types of white blood cells involved in the body's immune response.7 But questions of etiology and epidemiology were considerably more confusing. For one thing, it was already apparent that the "nationwide epidemic of immunodeficiency among male homosexuals"8 was in fact not restricted to gay men. According to the CDC's task force on the syndrome, 8 percent of the 159 cases were among heterosexuals, one of whom was a woman. In the pages of the New England Journal , Michael Gottlieb and his coauthors, the Los Angeles clinicians who had first reported the syndrome to the CDC, described finding the same syndrome in two exclusively heterosexual men, while Henry Masur and coauthors reported eleven cases of PCP in the New York area—five injection drug users, four gay men, and two men who were both.
Nonetheless, the focus of attention in all the medical literature remained squarely on the male homosexual sufferers, as evidenced by descriptors such as Brennan and Durack's "Gay Compromise Syndrome"9 and Masur et al.'s more euphemistic "Community-Acquired Pneumocystis Carinii Pneumonia."10 All speculation about causes proceeded from the premise of the centrality of male homosexuality. In Durack's words: "What clue does the link with homosexuality provide? Homosexual men, especially those who have many partners, are more likely than the general population to contract sexually transmitted diseases. Lesbians are not, and this apparent freedom, whatever its explanation, seems to extend to Kaposi's sarcoma and opportunisticinfections." Yet the assumption that the syndrome was somehow linked with homosexuality actually did little to immediately clarify the etiology, as Durack and others realized. Noting that "male homosexuals are at increased risk for the acquisition of common viral infections" such as cytomegalovirus (CMV), hepatitis B, and Epstein-Barr virus, Durack described the "obvious problem" with the hypothesis that CMV, or any of these viruses, might be the cause: "It does not explain why this syndrome is apparently new. Homosexuality is at least as old as history, and cytomegalovirus is presumably not a new pathogen. Were the homosexual contemporaries of Plato, Michelangelo, and Oscar Wilde subject to the risk of dying from opportunistic infections?"11 Durack's supposition was that "some new factor," such as poppers, "may have distorted the host-parasite relatinship." Concluding with some "frank speculation," Durack put forward a model essentially identical to the one Mildvan had proposed to the Native: that "the combined effects of persistent viral infection plus an adjuvant drug cause immunosuppression in some genetically predisposed men."
This model, which was sometimes called the "immune overload" or "antigen overload" hypothesis, represented the initial medical frame for understanding the epidemic: the syndrome was essentially linked to gay men, specifically to the "excesses" of the "homosexual lifestyle." The epidemic coincided historically, Newsweek suggested in the article "Diseases That Plague Gays," "with the burgeoning of bathhouses, gay bars and bookstores in major cities where homosexual men meet."12 Urban gay men, enjoying "life in the fast lane," had subjected themselves to so many sexually transmitted diseases, taken so many strong treatments to fight those diseases, and done so many recreational drugs that their immune systems had ultimately given up altogether, leaving their bodies open to the onslaught of a range of opportunistic infections. As one Harvard doctor is reported to have put it informally, "overindulgence in sex and drugs" and "the New York City lifestyle" were the culprits.13 What distinguished gay men from CMV-infected, sexually adventurous heterosexuals, and from cardiac patients inhaling amyl nitrite, and from the many patients who took strong antibiotic or antiparisitic drugs was, these experts suggested, that only gay men (or those gay men living in the "fast lane") confronted all these risks at once.
The Politics of Lifestyle
The speculative focus on "the gay lifestyle" casts light on the very nature of epidemiological science. When a mysterious illness appears in a specific social group, it makes eminent sense to ask what distinguishes that group from others not affected, or less affected, by the illness. The difficulty is that the isolation of "difference" presupposes a common understanding of what constitutes the "background" against which this difference stands out. In this sense, epidemiology is inevitably a "normalizing" science, employing—and reinforcing—unexamined notions of normality to measure and classify deviations from the norm.14 Faced with a "gay disease," epidemiologists immediately fastened upon the most sensational markers of homosexual difference, trumpeting the cases of men with histories of thousands of sexual partners, while ignoring the cases, also reported by clinicians from the very beginning, of gay men who were monogamous or who engaged in relatively modest amounts of sexual experimentation.15
With the advantage of hindsight, it is easy to recognize that the initial link between gay men and the new syndrome—while certainly the single most consequential aspect of the social construction of the epidemic—in fact reflected the confounding influences of what Irving Zola has called the "pathways" from doctor to patient. As Zola concluded from a more general study, it is often the case that apparent epidemiological differences in the incidence of some medical conditions actually derive from "factors of selectivity and attention which get people and their episodes into medical statistics..."16 There are important reasons why people do or do not seek medical help, yet, as Eliot Freidson has noted, the doctor tends to "assume that the cases he sees are no different from those he does not. And so he develops conceptions of illness that may have an inaccurate and artificial relationship to the world."17
To put it simply, some people get better medical attention, which means that medical professionals "attend" to their "unique" conditions. In New York City, if not elsewhere, it appears likely that there were at least as many cases of pneumocystis pneumonia among injection drug users as among gay men at the time of he discovery of the syndrome.18 But gay men, some of them affluent and relatively privileged, found their way into private doctors' offices and prominent teaching hospitals—and from there into the pages of medicaljournals—while drug users often sickened and died with little fanfare. Even as cases among injection drug users began to be reported, the "gay disease" frame for understanding the epidemic was already falling into place. Colloquially, the epidemic became known among some medical professionals and researchers in early 1982 as "GRID": Gay-Related Immune Deficiency.19
The power of frames as organizers of experience is precisely that they work to exclude alternative ways of interpreting an experience.20 Because "GRID" was a "gay disease," medical practitioners and researchers sometimes resisted the idea that it might appear elsewhere, and those who proposed that the epidemic could affect other people risked being discredited within the scientific community.21 Randy Shilts described how, throughout 1981, "there was a reluctance [at the CDC] to believe that intravenous drug users might be wrapped into this epidemic, and the New York physicians also seemed obsessed with the gay angle.…" "He says he's not homosexual, but he must be," doctors would confide to one another. One New York pediatrician was rediculed for his contention as early as 1981 that he was seeing children suffering from the same immune dysfunction as homosexual patients.22
But the differences in access to health care and the accident of the initial discovery of the syndrome among gay men are not adequate to account for the potency of this frame or the ease with which it fell into place. If gay men were perceived as plausible victims of a medical syndrome, it was in part because in the medical literature their sexualized lifestyle was already depicted as medically problematic. On one hand, epidemiologists and clinicians were genuinely surprised by the appearance of such devastating illness among "previously healthy" homosexual men. On the other hand, they were quick to make use of the existing stock of medical knowledge linking gay men with disease—specifically, the literature on sexually transmitted diseases among gay men that was published in the years just prior to, or coincident with, the onset of the epidemic.23 This literature, which was often cited by early medical claims-makers discussing the new epidemic of immune dysfunction,24 described an explosion of venereal diseases among gay men, the apparent aftermath of the "sexual revolution" and gay liberation. Concluding that homosexuality must be considered a risk factor in infectious disease, these articles stressed the need for clinicians to confront what one referred to as "homosexual hazards."25
Of course, modern conceptions of gay identity have always beenpartially medicalized. The very term "homosexual" dates from the nineteenth-century literature of doctors and sexologists. Gay identities have been formed, over the past one hundred years, through the dialectical interplay between an affirmative process of self-definition by homosexuals and the imposition of models by various groups of expert claims-makers.26 In this sense, what is ironic about the medicalization of gay male sexuality in the years just prior to the gay beginning of the epidemic is that is presupposed the successes of the gay movement, which were in part directed against an earlier medicalization (or "psychiatrization"). In opposition to a conception of homosexuality as a "mental illness," gay activists had put forward a positive conception of gay identity and the gay community. And in fact, the new medical discourse on gay men took as its starting point a particularly recent conception of the "lifestyle" of the urban gay male; this discourse marked the entry of the modern "gay community" into medical history.
While it cannot be doubted that doctors were genuinely concerned with treating the venereal diseases of gay men, the issue was framed in particular ways that influenced medical perceptions of homosexuality. First, the key phrases that were used—"homosexual hazards," "gay bowel syndrome," "homosexuality as a risk factor"—posed the problem essentially as one of identity and "lifestyle," rather thatn contraction of specific infections.27 (It seems far less likely that any medical journal would refer to "heterosexual hazards.") Second, the use of abstract, universalizing terms such as "the gay way of life" masked the considerable diversity of the life experiences and sexual practices of gay men;28 such stereotypes obscured the fact that researchers had made no attempt even to define, let alone systematically sample, the communities they characterized with rather sweeping generalizations.29
Yet this was the understanding of gay male sexuality that informed medical speculation in the early days of AIDS. To be sure, the reasoning behind the immune overload hypothesis was not irrational, and the hypothesis was not absurd: after all, the epidemic was being observed mainly among gay men; many of these men did have many sexual partners; many sexually active gay men were known to contract sexually transmitted diseases, as well as use poppers and other drugs. But the strength of the resulting hypothesis depended on a long chain of implicit assumptions—that the syndrome was in essence linked to homosexuals (and the cases among heterosexuals could be explainedaway); that the link to gay men meant that the epidemic was related to gay men's sexuality; that if gay men (by this view) were "promiscuous," then the illness must be a consequence of their promiscuity; and crucially, that repeated exposure to sexually transmitted pathogens (and to drugs) was actually capable of causing the immune system damage being observed. Furthermore, there was the assumption that the recent reported increases in rates of sexually transmitted disease and of drug use among gay men were indeed of sufficient magnitude to explain why the syndrome was emerging when it was.
As an initial hypothesis, immune overload was probably no more or less reasonable than many in the history of epidemiology or medical science. Nor was it ever hegemonic. For example, the first editorial on the syndrome in Lancet, the influential British medical journal, speculated on everything from "new or unrecognised environmental pollutants" to "even another infective agent";30 and such conjecture continued in the medical and scientific literature throughout 1982 and 1983. where "immune overload" (or, more generally, what Murray and Payne call the "promiscuity paradigm")31 exerted its greatest influence was outside the world of mainstream scientific practice. Reinforced by the mainstream media and filtering out into diverse arenas, the idea of a linkage between homosexuality, promiscuity, and illness informed an emergent sensibility about the syndrome—a vision, sometimes an unarticulated perception, of the epidemic as somehow the product of "the homosexual lifestyle." At times it has been voiced as a direct accusation: as late as October 1987, North Carolina's Jesse Helms could stand on the U.S. Senate floor and proclaim that "every case of AIDS can be traced back to a homosexual act."32 The notion that gays brought on the AIDS epidemic—and should be held responsible for having done so—has persisted long after the decline of mainstream biomedical support for etiological arguments focusing on "the gay lifestyle."
The idea that homosexuality "causes" AIDS also indicates the tangle of meanings packed into one short word—"cause"—and the difficulties involved in carrying out a conversation about causation that cuts across a range of lay and specialist communities. As Jana Armstrong has observed, "the word 'cause' is embedded in the language of public policy, the language of cell biology, the language of epidemiology. But the word does not mean the same thing in every instance of its use."33 Even a glance at a medical dictionary generates confusion: one such dictionary distinguishes between constitutional causes, exciting causes, immediate or precipitating causes, local causes, predisposingcauses, primary causes, proximate causes, remote causes, secondary causes, specific causes, and ultimate causes.34 ("Etiology" fares little better, since the definition of that term points back to the word "cause.")
Generally speaking, medical doctors were interested in finding a "primary cause"; that is, "the principal factor contributing to the production of a specific result"—in this case, the destruction of cell-mediated immune responses. But epidemiologists, in their focus on identifying risk groups, were in effect concerned significantly with "predisposing causes": "anything that renders a person more liable to a specific condition without actually producing it." Outside of the medical and scientific professions, the various usages of the word "cause" not only blurred these meanings but embedded notions of causation within a more general vocabulary of moral blame. Like cholera epidemics, which in the nineteenth-century United States were blamed on the squalid lifestyle of the poor;35 like gonorrhea, once regarded even by doctors as arising "from the continual irritation and excitement of the generative organs" of prostitutes;36 like smallpox and leprosy, which were blamed on the "unclean" practices of the U.S. Chinese population in the late nineteenth century;37 the genesis of the new epidemic of immune dysfunction was considered all too often with a view to assigning culpability. Partly through the power of the medical definitional process, partly through the ideological work of the opponents of gay liberation, gay men increasingly came to be equated with the emergent epidemic—it came to constitute part of their social identity.
Claiming the Epidemic
How did members of the affected communities respond to these formulations? Gay communities in the United States were both contributors to the "gay disease" frame and important critics of it. Initially, rumors of various lifestyle risks—a microbe in the water supply or the ventilation system at the most popular bathhouses, for example—spread rapidly through gay communities.38 Writers in the gay press showed little tendency, early on, to dispute the homosexual connection, as evidenced by the frequent use of the locution "gay cancer" (though often in quotation marks) to characterize the epidemic in 1981. This phrasing, however imprecise, effectively served as a rallying cry to alert gay men to the presence of a new danger.
Since many of the early reports in medical journals were written byclinicians well connected to gay communities, who were treating the patients in question, many gay people—and particularly health writers such as Lawrence Mass, himself a physician—were inclined toward sympathetic views of the medical and public health authorities. Increasingly, however, many gay writers, especially in the more left-leaning publications, were openly critical of medical researchers' tendency to blame the epidemic on gay promiscuity.39 Much as an earlier generation of feminists had conceived of medicine as a sexist institution, these writers and activists argued that medical science was a heterosexist and sex-phobic institution that reinforced norms of sexual conformity.
Gay physicians, such as those who were members of Physicians for Human Rights, an organization of gay doctors, found themselves at the felcrum.40 On one hand, they were called on to introduce their professional colleagues and epidemiological investigators to many specific aspects of the "gay lifestyle," often running up against a judgmental reception within the biomedical establishment. On the other hand, they felt a sense of responsibility to warn their communities about suspected risk behaviors—but knew they would lose credibility if they were perceived to be "sex-negative" or puritanical, given that gay liberation as a political movement was so closely tied to sexual liberation as a personal ethic.41
By early 1982, gay and lesbian activists had created two grassroots organizations that would prove to be pivotal in confronting the epidemic: the Gay Men's Health Crisis in New York City and, across the country, the Kaposi's Sarcoma Research and Education Foundation, later renamed the San Francisco AIDS Foundation. A testament to the high degree of political mobilization and access to resources in gay communitiess at the time, the appearance of these organizations marked simultaneous attempts to provide services to people suffering from the syndrome, relay relevant information rapidly to gay men at risk, and serve as an organized voice regarding questions of public policy.42
In the early period, these organizations took no position on the question of etiology. As the Gay Men's Health Crisis advised gay New Yorkers in an open letter in mid-1982: "Unsettling though it is, no evidence exists to incriminate any activity, drug, place of residence or any other factor, conclusively, in the outbreak facing us."43 At the same time, simply by organizing gay communities to confront—and, in effect, claim—the epidemic, these organizations helped to solidify the popular connection between the syndrome and homosexuality (aseven the name "Gay Men's Health Crisis" implied). AIDS became a "gay disease" primarily because clinicians, epidemiologists, and reporters perceived it through that filter, but secondarily because gay communities were obliged to make it their own.
Lifestyle vs. Virus (1982–1983)
The Expansion of Risk
Sensitive to the fact that gay doctors and activists criticized the informal "GRID" designation, the CDC came up with an official name for the epidemic in May 1982 and first used the term in print in September of that year.44 This name was chosen specifically for its neutrality—Acquired Immunodeficiency Syndrome, or AIDS: "acquired" to distinguish it from congenital defects of the immune system; "immunodeficiency" to describe the underlying problem, the deterioration of immune system functioning (and specifically, a decline in the numbers of helper T cells, causing the body to lose most of its capacity to ward off infection); and "syndrome" to indicate that the condition was not a disease in itself, but rather was marked by the presence of some other, relatively uncommon disease or infection (like PCP or Kaposi's sarcoma), "occurring in a person with no known cause for diminished resistance to that disease." This was strictly a "surveillance" definition, for epidemiological reporting purposes: it did not imply any knowledge about what AIDS "really was." But in the absence of a lab test for a known cause, this definition at least allowed the CDC a crude measure of the scope of the problem.
The newly defined syndrome would, over the course of 1983, achieve the status of a "Worldwide Health Problem," as the headline of one of Lawrence Altman's articles in the New York Times labeled it in November. By that time, AIDS cases would be reported "in 33 countries and all inhabited continents."45 Though most cases were in the United States or Europe, the most striking aspect of the epidemic's spread was the discovery of AIDS in equitorial Africa. In April the Washington Post summarized reports appearing in both Lancet and the New England Journal that described cases of AIDS in European countries, but among patients who had immigrated from or traveled in countries such as Zaire and Chad. Of twenty-nine such cases in France, six patients had become ill before June 1981—that is, before the epidemic was first reported in the United States.46 Immediatelyscientists and reporters in the West picked up on the notion that Africa "could have been the breeding place" for the epidemic.47
Despite the globalization of the epidemic and the formal change in terminology, the "gay disease" formulation, in various guises, continued to undergird medical investigation of the syndrome through the first half of 1982. For example, an editorial in the Annals of Internal Medicine by Dr. Anthony Fauci, a distinguished scientist who would later become the head of the AIDS program at the National Institutes of Health (NIH), laid out a number of etiological possibilities: "Is there a new virus or other infectious agent that has expressed itself first among the male homosexual community because of the unusual exposure potential within this group? Is this an immunosuppressed state due to chronic exposure to a recognized virus or viruses? Is this illness due to a synergy among various factors such as infectious agents, recreational drugs, therapeutic agents administered for diseases that are peculiar to this population such as the 'gay bowel syndrome' …?" But what Fauci never doubted was that the "critical questions" were: "why homosexual men and why occurrence or recognition only as recently as 1979?"48
Suddenly, this whole framework for understanding the epidemic was dramatically challenged. On July 9, the CDC reported thirty-four cases of Kaposi's sarcoma or opportunistic infections among Haitians living in five different states in the United States. None of those interviewed reported homosexual activity, and only one gave a history of injection drug use.49 The following week the agency reported three cases of PCP in people with hemophilia, all of them recipients of a blood product called Factor VIII, "manufactured from plasma pools collected from as many as a thousand or more donors."50 The CDC refrained from drawing conclusions, but noted that the occurrence of the hemophilia cases "suggests the possible transmission of an agent through blood products." Since bacteria were screened out of Factor VIII in the production process, while smaller particles such as viruses could potentially escape the screen, the "agent" in question would almost certainly have to be a virus.
Mass, writing in the Native, quickly noted the significance of these findings: of all the existing etiological hypotheses, "only that of viruses would seem able to provide a unitary hypothesis that could explain the sudden appearance of AID [the Native's term at that time] in a growing number of distinct populations." But he also acknowledged the alternative possibility: Perhaps, he suggested, "we are dealing witha number of superficially similar epidemics, each with its own primary etiology."51
Germs and Magic Bullets
One syndrome, one caise; many syndromes, many causes: these options suggested not just different etiological hypotheses, but opposing theoretical approaches to the understanding of human illness. Indeed, one of the most intriguing aspects of the early popularity of the immune overload hypothesis was that so many clinicians would readily forsake the approach to disease causation frequently described as the cornerstone of contemporary biomedicine: the principle of "one disease, one cause, one cure." As Allan Brandt has expressed it: "In this paradigm, individuals become infected with a parasite that causes dysfunction of some sort; disease is defined as a deviation from a biological norm. Social conditions, environmental phenomena, and other variables are generally discounted as causes of disease. The physician dispenses 'magic bullets' that restore the patient to health."52
Ever since the bacteriological revolution of the late nineteenth century, when germs replaced "miasmas" as the preferred explanation for illness, medical research typically has focused on the discovery of discrete microbial causes for specific diseases. To be more precise, two separate assumptions have been welded together: that most illnesses have a single, fundamental cause, rather than multiple necessary causes; and that the search for the cause of illness should focus primarily on microbes, very secondarily on lifestyle issues, and only incidentally on environmental causes related to the larger organization of the society. Of course, the monocausal/microbial approach has always had its critics. Writing in 1959, René Dubos characterized the "doctrine of specific etiology" as "unquestionably the most constructive force in medical research for almost a century," but noted that "few are the cases in which it has provided a complete account of the causation of disease." Citing the failures, "despite frantic efforts," to find cures for diseases such as cancer and mental illnesses, Dubos argued that the "search for the cause may be a hopeless pursuit because most disease states are the indirect outcome of a constellation of circumstances. …"53
Modern-day epidemiologists, more open to multicausal approaches to disease, may speak of a "web of causation" or may endorse "ecological" and "synergistic" models of illness that emphasize thecomplex interrelationships among environmental and host factors.54 But in laboratories, examining rooms, and medical school classrooms, the doctrine of specific etiology holds sway. Many analysts have seen in the monocausal/microbial model of disease the clue to medicine's ideological function within a capitalist society: it encourages people to attribute their illnesses to invisible particles rather than to occupational hazards or defects of social organization.55 But to understand why clinicians and researchers themselves reach for such explanations, the suggestions of sociologist Andrew Abbott may be more to the point.
First, the germ theory of disease focuses public attention on medicine's greatest triumphs and away from arthritis, heart disease, cancer, and other chronic problems that have proven less amenable to therapeutic success. Second, one of the chief legitimating values of medicine (like other professions) is its perceived efficiency; and as opposed to environmental explanations, monocausal/microbial ones lend themselves to neat and straightforward solutions ("Take two pills every four hours").56 The search for microbes enhances the power of laboratory researchers, who alone have the tools to conduct it. The search for environmental causes is, by contrast, frequently beyond their ken. In short, the commonly expressed preference of clinicians and biomedical researchers for simple, monocausal, microbial models may in an immediate sense have less to do with medicine's role in legitimating society than with doctors' and scientists' roles in legitimating scientific medicine.
In light of the prevailing explanatory preference, early clinical fascination with "the homosexual lifestyle" is all the more noteworthy. Of course, perceptions that illness is linked to lifestyle have become more common in recent years, with increasing attention to the relation between such factors as stress or eating habits and the development of various diseases. But to the extent that doctors endorsed a multicausal lifestyle model, they were going against the prevailing medical currents. It is well worth asking whether they would have been as likely to do so had it not been for the perception that "the gay lifestyle" was peculiarly laden with a potential for medical hazard.
Dissent at the Front Lines
Once put squarely on the table, the notion of a single, unifying cause of AIDS carried with it immediate practical implications.
The virus theory, now described by the Los Angeles Times as "a potentially much more serious candidate for the cause," was also a scary one, as it "raised the specter" of a communicable disease that might potentially affect anyone.57 Or as Newsweek warned in August, "the 'homosexual plague" has started spilling over into the general population."58 With the news, toward the end of 1982, of a case of AIDS having developed in a blood transfusion recipient twenty months old (one of whose donors was found to have AIDS)59 and of other cases in the female sexual partners of intravenous drug users with AIDS,60 the viral hypothesis gained increasing credibility. It wasn't that the lifestyle theories were immediately abandoned: for example, a news report in the influential Journal of the American Medical Association (JAMA ) surveying the controversy in late September 1982 claimed straightforwardly that "it seems unlikely that a virus alone is inducing AIDS," and devoted significant attention to the poppers theory and others, as well as to two researchers' discovery of a genetic marker in patients with Kaposi's sarcoma.61 However, the greater the number of risk groups, the less relevant seemed the details of "lifestyle," and the more attractive the notion of a unifying cause that could account plausibly for all manifestations of the syndrome.
In response to this challenge, advocates of the immune overload hypothesis struggled to fortify their claims. A central player was Dr. Joseph Sonnabend, a South African-born physician and researcher. A specialist in infectious diseases, Sonnabend had also done research in England on the drug interferon, put in a stint at Mt. Sinai School of Medicine as an assistant professor of microbiology, and served as the director of venereal disease control for the New York City Department of Health. At the time the epidemic emerged, Sonnabend was practicing as a community doctor in New York's Greenwich Village, largely treating the sexually transmitted diseases of his many gay male patients.62 Sonnabend had little inclination toward simple, monocausal models of illness. As he would later put it, his South African medical education had stressed that "if you want to understand sickness, you have to understand the environment in which sickness occurs."63
Writing in the Native in September 1982, Sonnabend sought support for his views by warning his readers that endorsement of the viral hypothesis could result in antigay discrimination and prejudice: "To publicly propose that any minority group carries a specific infectious agent capable of causing severe immunodeficiency and cancer is an act of tremendous seriousness. Given the potential repercussions, it vergeson the irresponsible that the suggestion is made on the basis of evidence that remains conjectural."64 The "conjecture," in Sonnabend's view, rested on twin assumptions, both of them debatable: that the same disease was present in "at least four disparate groups" (Haitians, intravenous drug users, hemophiliacs, and gay men) and that the disease was actually new in each group. While Sonnabend was careful to make clear that he was not ruling out these possibilities, his central thrust was to question them by suggesting that there might be various syndromes of immunodeficiency, all with similar symptoms. But rather than attempt to explain the cases of AIDS in the "newer" risk groups, Sonnabend focused attention on immunodeficiency in gay men. He proposed a variant of the original immune overload hypothesis, involving repeated infection with CMV and the reactivation of infection with Epstein-Barr virus. The bottom line, for Sonnabend, was summarized by the title of the Native article "Promiscuity Is Bad for Your Health." "This is not a moralistic judgment," Sonnabend insisted, "but a clear statement of the devastating effects of repeated infections."
As a neighborhood doctor, Sonnabend had limited resources with which to establish credibility, so his first step was to recruit allies.65 His advantage, as he later saw it, was his location on the front lines: "I was in the situation, as a physician to many of the men who actually developed the disease, to observe the guys in their setting. …"66 But his problem was that, within the research establishment, he was isolated:67 "I had this very unique access to information. But it was just myself, a lonely business. There were no takers, because I didn't have an important position in some medical center.… However, I did contact David Purtilo, the chairman of pathology at the University of Nebraska. Purtilo was known to me as an expert on Epstein-Barr virus."68
With Purtilo and another researcher, Steven Witkin, Sonnabend published an article in JAMA in May 1983, presenting his claims to a wide medical audience. (This article was later reprinted in the New York Academy of Science's 1984 volume on AIDS, as well as in JAMA 's "official" 1986 anthology of the epidemic, AIDS: From the Beginning .)69 Challenging the "prevailing view" that a novel infectious agent caused AIDS, the authors reminded their readers of "another acquired immunodeficiency syndrome," called Neapolitan disease, which "resulted from malnutrition and various viral infections." Similarly, the authors contended, multiple factors that included "recurrentcytomegalovirus (CMV) infections and immune responses to sperm are likely major causative factors" among promiscuous homosexual men. But on the crucial question of other "risk groups," the authors had little to say: "We cannot, at this time, explain why AIDS is thought to be occurring in Haitians, hemophiliacs, and others. Acquired immunodeficiency has many causes, including malnutrition, hormonal alterations, use of opiates and other intravenous drugs, and acute viral infections."70
Also in early 1983, Sonnabend became the editor of a new medical journal, AIDS Research , and declared its first issue "an appropriate occasion to review an alternative hypothesis regarding the genesis of AIDS." Here Sonnabend devoted more attention to the other "risk groups," noting that "transfusions are themselves immunosuppressive," since they expose the recipient to a variety of antigens in the donated blood, and that many Haitians are subject to tropical infections, as well as to poverty and malnutrition, both of which are highly correlated with illness. Sonnabend argued that "the risk groups are too broadly defined," and that focused epidemiological research was needed to tease out the specific risk factors (such as exposure to sexually transmitted pathogens in the case of gay men) actually linked to AIDS. Epidemiologists could also assess his model, he suggested, by comparing the prevalence of CMV in populations with and without AIDS.71
But Sonnabend had no epidemiologists at the ready, and in the absence of data, his hypothesis was, at best, informed speculation. Much the same could be said about all the "immune overload" theorists. Two years into the epidemic, there was little specific evidence in support of the claim that immune overload caused AIDS in gay men, let alone in other risk groups. Some research had reported on the health effects of poppers, including one study showing that of eight outwardly healthy gay men with signs of T-cell abnormalities, most were consumers of amyl nitrite.72 But the numbers were too small to warrant strong conclusions. More generally, epidemiologists and statisticians found it "exceedingly difficult to disentangle nitrite use from such other risk factors as the frequency of sexual encounters and the multiplicity of sexual partners."73 While the interwoven nature of these epidemiologically correlated behaviors could be taken as evidence of an overload model, it could just as easily reflect spurious associations. As Mass pointed out in a discussion of the limits of epidemiological thinking, "On the superficial basis of numbers alone …wearing handkerchiefed Levi's and having Judy Garland records in one's collection might also seem risky."74
Medical Uncertainty and Gay Skepticism
As the number of AIDS cases continued to rise, increasingly fearful gay men struggled to make sense of the shifting and indeterminate medical claims and to sort out the implications for their everyday lives. Toward the end of 1982, Gay Community News (GCN ), a left-leaning weekly based in Boston, cited the "growing consensus among experts that AIDS is transmissible … and most likely through sexual contact."75 Expressing the paper's sex-positive philosophy, writers in GCN were suspicious of views that blamed gay men for becoming sick, whether voiced by medical authorities or by gay men themselves. One prominent political analyst, Michael Bronski, wrote disapprovingly of the president of Gay Men's Health Crisis, quoting him as saying: "Something we have done to our bodies—and we still don't know what it is—has brought us closer to death."76 Another writer in the same issue of GCN applauded the views of Jim Geary, head of the Shanti Project, a San Francisco organization: "The reason [we] get sexually transmitted diseases is not because we have multiple sexual partners. … It's because we don't take the necessary precautions in having sex."77
The question of how to translate etiological uncertainty into guidelines for personal safety was deeply troubling to gay communities across the United States. Nowhere did the debate rage more fiercely than in the pages of the Native in late 1982. The Native 's editor and publisher, Chuck Ortleb, introduced side-by-side commentaries by stressing the paper's democratic impulse and the need for the general public to assess scientific and public health debates: "The articles printed on these pages provide good examples of the level of debate prevailing in medical circles. … Confusing? Contradictory? Of course. But then, so is much of the discussion surrounding the present health crisis. It's a discussion that we feel virtually everyone should be involved in—gay people as well as non-gay, laymen as well as physicians, policy-makers as well as the citizenry. …"78
One view in the debate was offered by Peter Seitzman, president of New York Physicians for Human Rights, the gay doctors' association.79 His argument was straightforward, if not altogether reassuring:"The available evidence overwhelmingly suggests that AIDS is caused by some as yet undiscovered transmissable [sic ] agent," probably a virus. Since the transmission pattern appeared to be similar to that of the hepatitis B virus, prevention guidelines would be "precisely the same as those for avoiding hepatitis B," namely, never use a syringe used by someone else and reduce promiscuity. Eager to avoid any implication of antisex attitudes, Seitzman reassured his readers that he himself had been "no more of an angel than Mae West." But he concluded by affirming the virtues of "monogamy as a survival technique," declaring that promiscuity is not immoral "but simply dangerous."
The "opposing" commentary by Michael Callen and Richard Berkowitz in fact arrived at roughly similar, if far more forceful, conclusions, but began with radically different etiological premises. It was also quite different in tone. Callen and Berkowitz identified themselves as twenty-seven-year-old men, both "victims of AIDS," each with a history of having been "excessively promiscuous." Although the article didn't say it, they were also both patients of Dr. Sonnabend and they endorsed his immune overload hypothesis. (Callen would go on to become the most prominent "long-term survivor" of AIDS in the United States, a familiar figure at rallies and demonstrations; until his death in 1993, he remained an activist and an ally of Sonnabend's.) Entitled "We know Who We Are: Two Gay Men Declare War on Promiscuity," the article was nothing short of a manifesto: "Those of us who have lived a life of excessive promiscuity on the urban gay circuit of bathhouses, backrooms, balconies, sex clubs, meat racks, and tearooms know who we are. … Those of us who have been promiscuous have sat on the sidelines throughout this epidemic and by our silence have tacitly encouraged wild speculation about a new, mutant, Andromeda-strain virus. We have remained silent because we have been unwilling to accept responsibility for the role that our own excessiveness has played in our present health crisis. But, deep down, we know who we are and we know why we're sick."80
Turning to the medical evidence in favor of the different causal hypotheses—which they had evaluated on the basis of "personal experiences," their talks with researchers and doctors, and their "own readings in both the medical and the lay press"—Callen and Berkowitz argued that "AIDS is not 'spreading' the way one would expect a single-viral epidemic to spread." The confinement of AIDS to specific risk groups, they maintained, should lead us to look for the specificexplanation for each group's immunosuppression. In practical terms, whatever theory one chose to believe, "the obvious and immediate solution to the present crisis is the end of urban male promiscuity as we know it today," the authors concluded. "The party that was the '70s is over," and anyone who defended promiscuity on political or ideological grounds was simply in denial.
Sharp responses to Callen and Berkowitz were quickly forthcoming, both in the Native and in other lesbian and gay publications. Charles Jurrist, writing "In Defense of Promiscuity," argued in his reply that the uncertainty of scientific knowledge about causation had to be factored into any evaluation of personal risk. While granting that the immune overload hypothesis "is the most plausible of the several theories concerning the origins of AIDS," Jurrist reminded readers that the hypothesis was far from proven: "It therefore seems a little premature to be calling for an end to sexual freedom in the name of physical health."81
Others, such as Michael Lynch, writing in the pages of the Canadian lesbian and gay newspaper The Body Politic , were even more assertive in defending sexual freedom against medical moralism. These writers put the issue in historical and political context: The medical critique of gay promiscuity was simply the latest of many attempts to portray gay sexuality as diseased. At stake in the debate was "gay identity" itself. Gay men had fought to construct an affirmative identity, an essential part of which involved strong defense of sexual freedom and a critique of puritanical attitudes. And many of those they had fought against were doctors and medical researchers. Given this history, it followed that the debate about gay men and their sexual practices was thoroughly interwined with an older power struggle: who was to say what it meant to be gay—the doctors, or gay men themselves?82 As Lynch poignantly expressed it: "Like helpless mice we have peremptorily, almost inexplicably, relinquished the one power we so long fought for in constructing our modern gay community: the power to determine our own gay identity. And to whom have we relinquished it? The very authority we wrested it from in a struggle that occupied us for more than a hundred years: the medical profession."83
In the absence of a cure for AIDS, or even an agreed upon cause—and in the aftermath of an initial scientific framing of AIDS as a "gay disease" linked to promiscuity, a formulation that aroused the wrath of many in gay and lesbian communities—the credibility of doctors,biomedical researchers, and public health authorities suffered greatly in those communities. Increasingly gays were prompted to respond by insisting on their own right to intervene—to weigh the evidence, pass judgment, and remind the medical establishment at every pass whose lives were really on the line. Gay doctors like Mass sought a moderate position; writing in the Native in response to Lynch's commentary in The Body Politic , he acknowledged: "To an enormous extent, what Lynch is saying is true. Mainstream medicine and psychiatry have in fact been largely responsible for contemporary stereotypes of homosexuals as 'abnormal,' 'perverse,' and 'sick.' At the same time, however, mainstream medicine and psychiatry continue to serve vital health needs." Maintaining that the problem was not with medical science but with "the political abuse of that science," Mass advised his readers to "be critical but remain open to well-qualified medical advice."84 But for those on either extreme of the promiscuity debate, from Lynch to Callen and Berkowitz, the watchword was self-reliance. Become your own expert: ultimately, that was the only reasonable hope gay people might have of surviving. "Rely on no single source for your information," exhorted Callen and Berkowitz: "not your doctor, not this newspaper, not the Gay Men's Health Crisis, not the Centers for Disease Control."85
Of course, it goes without saying that this strategy of collective empowerment presumed the existence of gay doctors, gay newspapers, and a Gay Men's Health Crisis. To be sure, the spreading disease would decimate the ranks of existing gay leadership. But ironically, the gay response to AIDS both presupposed and furthered the social development of lesbian and gay communities and their political clout—a process that Dennis Altman has called "legitimation through disaster."86 Gay men and lesbians had long confronted homophobic attitudes, antigay discrimination, and heterosexist presumption on a daily basis in the workplace, in religious settings, in encounters with family members, on television, and in the movies. But in organizing to meet these challenges, lesbians and gay men had developed political and social institutions that were poised to respond to the new threat when it erupted. Moreover, gay communities were dominated by white, middle-class men—people with influence in society and access to an array of social, cultural, and political resources. It's no surprise that gays were hotly debating the details of causation theories while intravenous drug users—often the poorest of the poor—sat on the sidelines: these were the realities of power in the United States inthe 1980s. Even people with hemophilia, a diverse group that had the benefit of a preexisting national lobby, did not mobilize forcefully in response to the emergence of the epidemic. In this early period, Haitians were the only other group to challenge medical claims; they objected to the portrayal of Haiti as a possible origin of the epidemic and combated wild epidemiological speculation about the role of voodoo rituals in the transmission of AIDS. And most of the opposition came not from the grassroots but from politicians in Haiti and Haitian doctors living in the United States.87
The distinctiveness of gay communities' approaches to the emergence of the epidemic is brought out in Cathy Cohen's comparative analysis of gay and African-American responses. Although gay communities were hit harder, both gays and African-Americans came to be disproportionately represented in the statistics of illness and death as the epidemic proceeded. Both of these social groups were marginalized, and historical memory inclined both of them to distrust federal biomedical institutions. Yet "the indigenous norms and structures of these communities" promoted different political outcomes: "In the Gay community AIDS has become associated with the community's struggle for rights and entitlements. In the Black community much of the response to AIDS is based on a framing of the disease that still emphasizes behavior and the individual actions of those who have AIDS, making it much more difficult to transform AIDS into a political issue for the community."88 As Cohen has described, established African-American organizations eschewed "ownership" of the problem of AIDS. It was often up to black gays and lesbians—those who stood at the intersection of these two social groups—to try to mobilize African-American communities and start up new organizations, while simultaneously confronting racism within gay communities.89
The Triumph of Retrovirology (1982–1984)
Gallo's Family of Viruses
Writing in 1988, Robert Gallo, the NIH scientist who would share credit for discovery of what was to become known as the "human immunodeficiency virus," reflected back on the scientific effort to understand AIDS. Gallo concluded that progress had been made possible by "two general earlier developments": "first , by major advances that took place in basic sciences, particularly in immunology… and molecular biology …; and second , by the opening of the whole field of human retrovirology that, oddly enough, occurred only a few years before the AIDS epidemic. …" To this Gallo added, "There is no doubt in my mind that if AIDS came upon us full force in the 1960s or even in the early 1970s, we would still be wandering in the dark regarding most of what we know today."90
Gallo's "odd" fact can be put in different terms: At the moment when people began to suspect that AIDS might be caused by an infectious agent, there existed a small group of prominent scientist working in a very specialized area who were inclined to imagine that a "retrovirus" might be the cause, who were motivated to pursue that speculation, and who were well equipped to do so. Only in the 1960s had researchers discovered that the genetic material of certain viruses consists of RNA (ribonucleic acid) rather than DNA (deoxyribonucleic acid, often called the "blueprint of life"). Normally, viruses infect cells and turn them into virus factories, causing the cells to produce new viruses according to the specifications of the rival DNA. The virus's DNA is copied into RNA, which is then used to manufacture viral proteins; once the new viruses are assembled, they are ejected and go off to infect other cells. When viruses were found that consisted of RNA rather than DNA, they presented a puzzle to scientists, because it was unclear how the viruses could replicate. In 1970, however, in work that would win them the Nobel Prize, researchers David Baltimore at the Massachusetts Institute of Technology and Howard Temin at the University of Wisconsin independently discovered that these RNA viruses contained an enzyme, which they termed "reverse transcriptase," that copied the viral RNA into DNA. This DNA then served as the blueprint for the manufacture of new viruses. So while in normal viruses the sequence was "DNA to RNA to new viruses," in these unusual viruses there was an extra step: "RNA to DNA to RNA to new viruses." To describe the transcription from RNA to DNA and back again, virologists coined the term "retrovirus."91
During the "War on Cancer" in the 1970s, researchers such as Gallo at the NIH's National Cancer Institute (NCI) investigated links between animal retroviruses and various forms of cancer. However, until the late 1970s, no retroviruses were known to cause disease in humans. At that time, both Gallo and a group of Japanese researchers claimed credit for the discovery of one believed to cause adult T-cell leukemia, a rare form of cancer found mostly in Japanese fishing villages.92 Gallo named the retrovirus the human T-cell leukemia virus,or HTLV. This work earned him the Lasker Prize, the highest award in biomedicine short of the Nobel Prize. He found another virus in the same family in 1982, which he claimed caused a different type of leukemia; in consequence, the two viruses became known as HTLV-I and HTLV-II.
As Gallo described it in retrospect, when he first heard about the new syndrome in gay men in 1981, he had no reason to think it might be linked to a retrovirus and indeed had little interest in the issue.93 Or as Joan Fujimura has suggested more generally, scientists construct and pursue "do-able problems": they do not venture off in any direction at random, but rather structure their work by finding effective ways of integrating and coordinating the relationship between the experimental procedures at hand, the organization of their laboratories, and the social worlds through which they move.94 Given that the initial hypotheses focused on homosexual lifestyle risks, many virologists simply saw no particular reason to be interested.
By Gallo's account, his curiosity was piqued only in 1982, when James Curran of the CDC briefed NIH researchers about the epidemic, expressing to them his own belief that the syndrome was caused by an infectious agent, and stressing that one hallmark of the syndrome was the helper T-cell deficiency.95 This was enough for Gallo to hypothesize that the epidemic might be caused by HTLV or by a retrovirus of the HTLV family—by a close relative, that is, of the two viruses whose discovery had already brought him considerable acclaim within the world of virology. After all, HTLV specifically infected helper T cells; moreover, HTLV was known to be transmitted in blood and semen, which seemed also to be plausible transmission routes of the putative agent in AIDS. Finally, there was some precedent for a retroviral role in a condition like AIDS, since a feline retrovirus was linked to immune deficiency in cats. Gallo became convinced that AIDS was an HTLV-linked disease. Only some years later would he do an about-face and make an intriguing confession: "That hypothesis, as it turned out, was wrong. Nonetheless, it was fruitful, because it stimulated the search that led to the correct solution."96
Gallo had little patience for alternative hypotheses that were common at the time. The medical tendency to, as he put it, "round up the usual suspects"—CMV, Epstein-Barr virus, and the like—seemed to him unlikely to provide an explanation for what was, after all, a new epidemic. Nor was he impressed by the popular hypothesis of immune overload. Interestingly, he objected not just on empirical grounds—thatimmune overload seemed unlikely to account for manifestations in all the risk groups—but also on the basis of his understanding of causality in disease processes: "Whereas some complex diseases … are believed to involve different steps and sometimes different factors, most human disease (even some cancers) can be thought of as involving a primary causal factor. Certainly this has been the case for most past epidemic disease for which we in time did learn the cuase."97
Committed in general to what Dubos called the "doctrine of specific etiology," Gallo dedicated his laboratory to an investigation of his hypothesis: that AIDS was caused by the virus he was already working with and had invested in, HTLV. Within weeks of embarking on this search, Gallo's assistants found the leukemia virus in the T cells of two U.S. gay men, a Haitian woman who died of AIDS in France, and a Frenchman who had received a blood transfusion.98 Gallo sent two papers describing the findings to Science magazine, the preeminent general science publication in the United States. His colleague, Myron ("Max") Essex at the Harvard School of Public Health, sent along a third paper also reporting signs of HTLV infection in AIDS patients.
The French Virus
Meanwhile, across the Atlantic, a similar search for a retroviral cause of AIDS was proceeding according to different premises. In Paris, a group of physicians had been meeting informally to discuss the epidemic, and one of them, an immunologist named Jacques Leibowitch, who was familiar with Gallo's work on HTLV, had become convinced that a retrovirus was the cause. Skeptical of arguments about poppers and promiscuity,99 Leibowitch specifically hoped to demonstrate "that the cause of AIDS was not homosexually related."100 But neither Leibowitch nor any of his colleagues knew how to look for a retrovirus, so they set out to enlist the support of Luc Montagnier, chief of viral oncology at the famous Pasteur Institute, a private, nonprofit research institution founded by Louis Pasteur in 1887.
The physicians' group had a hunch that if a virus was causing the depletion of T cells, then there might be higher levels of virus present in people who were at an earlier stage of illness, before most of their T cells had been killed off. So they sent Montagnier samples of lymph tissue from a gay male patient with "lymphadenopathy syndrome"—acondition of chronically swollen lymph glands, increasingly prevalent among gay men and believed by many to be a precursor to AIDS. Montagnier's research team extracted T cells from the tissue and put them in an incubator with nutrients, hoping to grow a virus. When tests showed the presence of reverse transcriptase, the enzyme that is the distinctive marker of retroviruses, they knew they had found something. The reverse transcriptase activity rose and then fell—a sign that the virus was killing its host cells—but by adding fresh cells from new sources, the French researchers were able to maintain the culture. With the aid of electron microscopy, the Pasteur group also succeeded in photographing viral particles.
When Montagnier contacted Gallo in early 1982 and informed him of his findings, Gallo encouraged him to submit his paper to Science , so that Gallo's, Essex's, and Montagnier's papers could all appear together. Since Science allows its authors to suggest appropriate peer reviewers, Gallo told Montagnier he would be happy to review the Pasteur Institute;s findings for the magazine. In his comments to Science , Gallo urged rapid publication, stressing the importance of Montagnier's work. But in addition, as reported John Crewdson has described in a highly critical exposé of Gallo's work, Gallo offered to write the abstract, which Montagnier had neglected to include. Gallo's abstract identified the French virus as a "C-type retrovirus," similar to Gallo's HTLV.101 Gallo had effectively enlisted the Pasteur researchers behind his own HTLV.
The papers appeared in Science in May 1983,102 where, as Crewdson noted, they "made a considerable splash."103 But few people paid much attention to Montagnier's paper, which followed the other three in the pages of Science ; it appeared simply to confirm the findings of the American researchers. As Jay Levy, a virologist at the University of California at San Francisco Medical School (UCSF), who had also embarked on a search for a retroviral causative agent, later recalled, "The write-up in the Science papers sounded like the French virus and the Gallo virus were the same."104
But Montagnier and his collaborators suspected otherwise. Their photographs didn't especially resemble HTLV. And when they exposed their virus to HTLV antibodies, they didn't observe any "cross-reaction"—as they should have, if the virus were really a close cousin of HTLV. Most crucially, their virus killed T cells in the test tube. HTLV caused its host cells to multiply wildly—the hallmark of cancer. Of course, the French had no actual proof at this point that the virusthey had found was indeed the cause of AIDS. But they were increasingly convinced that theirs was a previously undiscovered retrovirus, not HTLV or even a member of the same family; and they set out to demonstrate its causal relationship to the epidemic.
By fall 1983, at the virology conference held each September in Cold Springs Harbor, New York, Montagnier could report finding his virus—which he was now calling "LAV," or lymphadenopathy-associated virus—in about 60 percent of patients with lymphadenopathy syndrome and 20 percent of those with AIDS. None of these patients appeared to be infected with HTLV. At the conference, Gallo angrily disputed Montagnier's findings, claiming that the French measurements had to be in error. (Much later, Gallo would write: "I have come increasingly to regret that the tone or spirit of my questioning that day was too aggressive and therefore misunderstood.")105 What Gallo did not mention to the conference-goers was that, despite his own lab's best efforts, he and his associates had been unsuccessful in finding HTLV in the majority of samples from AIDS patients that they had been studying over the past several months.106
An Addition to the Honor Roll?
The events of the subsequent seven months are obscure, and—despite intensive scrutiny by journalists and a half dozen official investigations by various reputable bodies—the facts may never fully be known. What appears beyond dispute is that, shortly after the Cold Springs Harbor conference, Montagnier forwarded to Gallo a sample of LAV for him to study. Then, the following April, reports of Gallo's discovery of a "third variant" of HTLV began to appear in the pages of U.S. newspapers. Just months after insisting that HTLV was the cause of AIDS—while increasingly having trouble finding it in AIDS patients—Gallo presented the world with a new virus, "HTLV-III," which he claimed was a member of the HTLV family. Later, in January 1985, investigators would determine that Gallo's HTLV-III samples had a 99 percent genetic similarity to Montagnier's LAV—that is, the viruses were much too similar to have come from separate sources. The implications were clear: Whether the consequence of accidental contamination of viral cultures—a common problem in virology labs—or of outright theft and misrepresentation, the Pasteur Institute's LAV had found its way into Gallo's cultures. Almost beyond a doubt, Gallo had in fact "discovered" Montagnier's virus.107 Yet none of this was known in 1984. Indeed, there was little said about Montagnier on April 23 that year, when Margaret Heckler, President Ronald Reagan's secretary of health and human services, stood before a roomful of reporters. "The probable cause of AIDS has been found," she announced with some fanfare: "a variant of a known human cancer virus, called HTLV-III." Just a few days earlier, Lawrence Altman, the New York Times 's medical reporter, had received a scoop from CDC Director James Mason, who told him that a virus discovered in France, called LAV, was the likely culprit; the Times had run the story on the front page.108 "There was so little excitement in the scientific community when the French came up with their announcement last May," noted Mason, claiming he did not understand why it had taken so long for the importance of the Pasteur Institute's work to be recognized. But at the press conference on Tuesday, Mason's boss had a different tune to play.
"Today we add another miracle to the long honor roll of American medicine and science," said Heckler. "Those who have said we weren't doing enough," she added, in response to widespread complaints of inactivity on AIDS by the Reagan administration, "have not understood how sound, solid, significant medical research proceeds." As Randy Shilts described it, the researchers on the podium with Heckler "blanched visibly when she proclaimed that … a vaccine would be ready for testing within two years."109
Heckler made only brief reference to the Pasteur Institute scientists, describing them as "working in collaboration with the National Cancer Institute"; she indicated her belief that LAV and HTLV-III "will prove to be the same."110 Nor was mention made of UCSF virologist Jay Levy, who was also close on the heels of a virus linked to AIDS. (He would submit his paper to Science the following month.)111 Pressed by puzzled reporters, Gallo added: "If it [the virus] turns out to be the same I certainly will say so…."112 Heckler emphasized the crucial role of the U.S. research, noting that only Gallo had succeeded in reproducing large quantities of the virus, which was necessary for the development of a blood test that could detect viral antibodies.113 Hours earlier, the U.S. government had filed a patent application for just such a test.114
The New York Times , in an editorial printed a few days afterward, was not slow to draw implications from the episode. "In the world of science, as among primitive societies, to be the namer of an object is to own it," the Times noted wryly, pointing to the dispute between"LAV" and "HTLV-III." Since the blood screening test was not yet in commercial operation and no vaccine had yet been produced, "what you are hearing is not yet a public benefit but a private competition—for fame, prizes, new research funds."115
"Strong Evidence of a Causative Involvement"
Certainly one of many unusual aspects of the whole affair was that Heckler's press conference was held before Gallo's findings had even been published in a peer-reviewed forum—normally a serious breach of professional scientific etiquette in itself. Those who wanted more substantial information about Gallo's claims had to wait until May 4, when four articles by Gallo's group appeared in the pages of Science . This too was extraordinary. As Gallo later commented, "Getting one paper in Science is a lot. Getting two is fantastic. Getting three was a record. We had four at one time."116
Gallo used these articles to put forward a series of interconnected claims: that he had found a new virus; that he had succeeded in mass-producing it; that the virus was related to HTLV; that antibodies to the virus could be detected in blood; and, most crucially, "that HTLV-III may be the primary cause of AIDS." In the second of the four papers, Gallo and his coauthors focused specifically on the etiological argument. After reviewing the evidence in favor of an infectious agent as the cause of the syndrome, Gallo reminded his readers that "we and others have suggested that specific human T-lymphotropic retroviruses (HTLV) cause AIDS."117 Gallo's wording was also significant: he had redefined HTLV, from "human T-cell leukemia virus" in 1983, to "human T-lymphotropic retroviruses " in 1984. While the original name denoted the relation between a single retrovirus and a specific form of cancer, the new name described a family of viruses more vaguely characterized as "T-lymphotropic," that is, having an affinity for T cells. Gallo had reinvented "HTLV" so as to more plausibly encompass his new virus as a relative of HTLV-I and HTLV-II.
Moving on to HTLV-III, Gallo described detecting the virus in, and isolating it from, "18 of 21 samples from patients with pre-AIDS [the so-called lymphadenopathy syndrome], three of four clinically normal mothers of juvenile AIDS patients, three of eight juvenile AIDS patients, 13 of 43 … adult AIDS patients with Kaposi's sarcoma, and 10 of 21 adult AIDS patients with opportunistic infections." Althoughideally one would expect to find a primary causative agent in every case of a disease, Gallo noted that "the incidence of virus isolation reported here probably underestimates its true incidence since many tissue specimens were not received or handled under what we now recognize as optimal conditions." In contrast with these findings that associated the virus with the expression of AIDS was the striking absence of HTLV-III in cases where AIDS was also absent. Out of 115 clinically normal heterosexual blood donors, not a single one showed signs of the virus. And out of 22 clinically normal gay male donors, only one tested positive for the virus, and that person developed AIDS within six months. These studies, Gallo and his coauthors concluded, "provide strong evidence of a causative involvement of the virus in AIDS."118
What in fact had Gallo established? Four years later, in response to challenges about whether the virus had been proven to cause AIDS, Gallo would maintain: "In my opinion all of the sufficient data was available at the time the cause was first announced in the spring of 1984."119 But this was a difficult position to sustain, at least on the basis of Gallo's published findings. Gallo had shown that, in specific small samples, laboratory signs indicating the presence of his virus were often correlated with the expression of AIDS at what were believed to be two different stages of disease progression ("pre-AIDS" and AIDS). Moreover, there were no such signs of virus in clinically normal people, suggesting that the virus or viruses had some special relationship to AIDS. But just because HTLV-III and LAV were often correlated with the syndrome, did that mean they were causing it? AIDS, after all, was a syndrome whose hallmark was the presence of a range of opportunistic infections; perhaps HTLV-III and LAV were viruses that were contracted by people who already had weakened immune systems. Gallo would have been in a better position to respond to this challenge if he had had more cases like that of the clinically healthy but infected gay man who later developed AIDS. But the other 21 of his 22 "clinically normal homosexual donors" all tested negative for the virus, so there was really no evidence that HTLV-III infection was a precursor to immune system damage. (Three out of 4 of the "clinically normal mothers of juvenile AIDS patients" tested positive for the virus, but these numbers were small, and Gallo did not report that any of the women had subsequently developed AIDS symptoms.)120
When asked ten years afterward whether he had been able, atthe time of publication, to rule out the possibility that HTLV-III was an opportunistic infection, Gallo acknowledged that he could not. But "the evidence was overwhelming in my mind," Gallo recalled. "Science is never 100 percent. It's not mathematics. You play not on hunches, but on data that becomes overwhelming in your mind. …"121 To make a credible claim for "strong evidence of a causative involvement," Gallo was in fact relying heavily on the plausibility of HTLV-III as a pathogenic agent, given what was known about the virus and about AIDS. At least in vitro, HTLV-III killed helper T cells. And the central manifestation of immune system damage in people with AIDS was precisely the low numbers of those same helper T cells. Still, at this point, little was known about the effects of HTLV-III in vivo.
Koch's Postulates and the Proof of Causation
Epidemiologists and biomedical researchers rely on a range of principles to establish causation in disease. However, since the acceptance of different versions and interpretations of these principles of causation has itself become one of the stakes in the controversy over the causation of AIDS, these principles cannot be independently invoked as neutral measures. The most well known causation criteria are called "Koch's postulates," named after Robert Koch, the nineteenth-century German microbiologist. The postulates consist of four steps that are easily stated. First, the causal agent must be found in all cases of the disease. Second, the agent must be isolated from a carrier and grown in pure culture. Third, when the culture is injected into a susceptible laboratory animal, the animal must contract the disease. Finally, the causal agent must then be recovered from the diseased animal.
The precise relevance of Koch's postulates to contemporary biomedical research (particularly with regard to viruses, which were unknown at the time of Koch's own work) is in dispute, and this dispute has been magnified as a result of recent debates about the causation of AIDS. Many have proposed that researchers nowadays must work with "modern" or revised versions of the postulates, and have argued that Koch himself did not intend them to be followed rigidly.122 Nevertheless, Koch's postulates remain a well-known reference point for considering questions of etiology in scientific medicine.
For instance, Richard Krause, the director of the National Institute of Allergy and Infectious Diseases of the NIH, gave a conference talk in the summer of 1983 on "Koch's Postulates and the Search for the AIDS Agent," noting that "technical difficulties" often "impede the fulfillment" of all of Koch's postulates, but concluding: "If we abide by the scientific guidance of Koch's postulates, we are sure to discover the cause of AIDS."123 Similarly, Lawrence Altman focused squarely on Koch's postulates in an article, published later in 1984, on "How AIDS Researchers Strive for Virus Proof."124 Altman presented Koch's postulates as an important medical tradition that researchers have looked to for a century, but he noted that doctors are sometimes forced to rely on immunological or other experimental evidence when Koch's postulates cannot fully be satisfied. With less equivocation, James D'Eramo wrote in the New York Native soon after the Heckler press conference: "The definitive classical proof that a virus or bacterium causes disease rests on causing the disease in animals by injecting them with the putative agent. AIDS has yet to occur in a laboratory animal."125 Dr. Nathan Fain, the medical writer for the national, West Coast-based gay newsmagazine the Advocate, made roughly the same claim in May when he explained why "work must continue to prove beyond all doubt that the candidate virus does cause AIDS."126
Clearly, if Koch's postulates are the benchmark, then Gallo's May 1984 articles in Science by no means established HTLV-III as the cause of AIDS. But since the criteria for proving causation have been contested, it may be useful to assess the credibility of Gallo's claims-making by looking at a relatively weak version of the causation criteria presented in a recent epidemiology textbook. According to Mausner and Kramer, the likelihood that an association is causal can be evaluated by examining several criteria.127 First, there is the strength of the association, which they describe as the "ratio of disease rates for those with and without the hypothesized causal factor": here Gallo's evidence is compelling but far from perfect, since he was able to isolate the virus only in fewer than half of the samples from people actually diagnosed with AIDS. Second, the "dose-response relationship": does a higher dose of the causal factor result in higher rates of disease expression? Gallo had no data on this point. Third, the consistency of the association across different studies: clearly this was yet to be determined. Finally, is the association a "temporally correct" one, meaning that the cause precedes the expression with a sufficient "induction period" or "latency period"? With the exception of theone virus-positive, clinically healthy gay man who developed AIDS within six months, Gallo had no relevant data to report.
Given the state of the evidence in early 1984, perhaps a more plausible claim was that of Jay Levy, whose results in isolating what he called "ARV," or AIDS-associated retrovirus, were published in August.128 Levy found signs of ARV in about half his AIDS patients and in about 20 percent of clinically healthy homosexual men, but in only 4 percent of clinically healthy heterosexuals. Levy recalled agonizing over how to phrase his conclusion: "I called a good friend of mine … who's an editor, and I said, 'How do I do this? I don't want to say it isn't, but I don't want to say it is .'"129 In the end, Levy's wording was cautious: "Although no conclusion can yet be made concerning their etiologic role in AIDS, their biologic properties and prevalence in AIDS patients certainly suggest that these retroviruses could cause this disease."130
But for Gallo, the notion that he had proven the virus to be the cause became something crucial to defend, particularly as his credibility on other claims was challenged. In 1985, the Pasteur Institute sued the U.S. government in a patent dispute over the discovery of the virus, and in 1987 the heads of the two governments, Jacques Chirac and Ronald Reagan, signed an agreement splitting the royalties for the commercial antibody test. Especially after it became apparent that Montagnier's LAV had found its way into Gallo's viral cultures—a point that Gallo would formally concede in 1991—Gallo gradually backed off from claiming any primacy. And although Gallo continued to present the discovery of HTLV-III as a natural outgrowth of HTLV research,131 he was eventually forced to accept the prevailing view that, from a genetic standpoint, the new virus was not reasonably classifiable as an HTLV virus. In response to the confusing array of acronyms then in use—HTLV-III, LAV, ARV, HTLV-III/LAV, and others—the Human Retrovirus Subcommittee of the International Committee on the Taxonomy of Viruses rebuffed Gallo and agreed on a new, compromise, name in 1986: HIV, human immunodeficiency virus. (Levy and Montagnier signed the agreement; Gallo and his close associate, Max Essex, dissented.)132
The Framing of AIDS
The naming of the virus by the Human Retrovirus Subcommittee marked the initial stabilization of "HIV" as a unitaryobject of medical knowledge.133 But even before this point, the illness AIDS had become a relatively stable cultural entity whose social meanings, however fluid and multiple, had at least begun to congeal. Over the course of a few years, AIDS had come to be "framed," or constructed, within the context of strong beliefs and attitudes about sexuality, promiscuity, and homosexuality and through recourse to a wide range of analogies: Was AIDS like cancer? Was it like herpes? Was it like hepatitis B? Was it an HTLV-like illness? AIDS itself had also come to serve as a frame for understanding other events and social behaviors. Perhaps most notably, as sociologist Steven Seidman has argued, "AIDS … provided a pretext to reinsert homosexuality within a symbolic drama of pollution and purity."134
These framings, and the associated stigma, had also provided possibilities for gay men to assert claims for "ownership" of the epidemic (however ambivalently), or at least some of the public responses to it.135 Indeed, the same social networks and institutional linkages that had permitted rapid amplification of a virus also gave rise to the organization of a concerted grassroots response. Lesbians, subject to what Erving Goffman has called a "courtesy stigma," or stigma by association, acted as collaborators with gay men in these efforts, often playing leadership roles.136 This extraordinary success of gay and lesbian communities in establishing their right to speak about the epidemic would fuel a willingness and capacity to challenge the knowledge-making practices of biomedicine in the coming years.
Biomedical researchers, and in particular virologists, had also staked out claims to "ownership" of AIDS, and had done so through powerful findings concerning a probable causal agent. The credibility of AIDS research would rapidly become linked to the credibility of this particular causal claim: between 1984 and 1986, the retroviral hypothesis would achieve near-hegemonic status among scientists. It would also, by and large, be taken for granted in the communities affected by the epidemic, in the mass media, and among the lay public. The pathways, mechanisms, and consequences of the "black-boxing" of HIV are the topics of the next chapter.
Excerpted from Impure Science by Steven Epstein Copyright © 1998 by Steven Epstein. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.