These were the clouds hanging overhead as thirteen thousand people from 166 countries assembled in Berlin's vast and labyrinthine conference center in June for the Ninth International Conference on AIDS. A cover story by Science reporter Jon Cohen had set the stage for the scientific proceedings. "The more we learn, the less certain we are" was the message Cohen had gleaned from a survey he conducted of 150 leading AIDS researchers: "After more than a decade of struggling in frustration as the epidemic gallops on, researchers are being forced to reexamine assumptions they once held without question." Even as "politicians and AIDS activists [were] demanding results immediately," researchers confronted a series of "collapsing certainties," including the virtues of antiretroviral therapy in asymptomatics, the trustworthiness of surrogate markers in evaluating
treatments, and the reliability of certain key vaccine experiments. In addition, "many researchers who once believed almost all the damage caused by HIV could be explained by the virus's direct killing of cells now think indirect mechanisms must also be at work."
Progress certainly was being made in the understanding of pathogenesis, as conference presentations by top scientists such as Fauci and Jay Levy made clear. On one hand, better laboratory techniques such as PCR had revealed higher concentrations of infected cells throughout the body (and not just in the bloodstream); this lent credence to claims that direct cell-killing could play an important role in AIDS, while arguing against those, such as Duesberg, who had been maintaining that HIV could not be the true cause of AIDS if so few cells were infected (see chapters 3 and 4). On the other hand, few now thought that direct cell-killing was the only factor at work. Levy's recent, one hundred-page journal article on pathogenesis had listed more than a dozen factors he believed were involved in HIV-induced immune deficiency. HIV and its constituent proteins appeared to have a range of effects on both infected and uninfected cells, as well as on their associated cytokines; aberrant signaling by the cytokines then caused cascading effects throughout the immune system.
Better understanding could be gleaned, perhaps, from the study of so-called "long-term survivors" or "long-term nonprogressors"—this was a direction that activists had been promoting and that was increasingly emphasized in Berlin. "There's a growing sense that there is no magic bullet for AIDS, so we should shift research to see why some people do well …," said ACT UP/New York's Aldyn McKean. But as Robin Weiss had pointed out in an article on pathogenesis in Science , there might be nothing "special" about long-term survivors: since "the rates of progression … are also consistent with a stochastic, random occurrence of AIDS after HIV infection …, 'long-term' survival could be pure luck."
Gallo, in his address, offered a dazzling array of future, high-tech treatment possibilities. Gene therapy and "antisense" therapy might be used to prevent infection of cells. Drugs could be developed that targeted cellular products used by the virus, rather than directly targeting the virus. HIV-infected people could be given a genetically engineered protein of HHV-7 (human herpes virus, number seven), a virus that "competes" with HIV for CD4 receptor sites; the protein would then beat the virus to the CD4 molecules on the T cells and prevent infection.
This was all very nice, but what treatments were available in the short run? What could the assembled physicians (or at least, those from rich countries) go back and tell their patients, symptomatic or asymptomatic, about the options that were open to them? What was the "state of the art"? One crucial set of facts was offered by the Concorde researchers, who presented a range of data that went beyond the limited report in Lancet . Overall, the Concorde trial gained in credibility at the conference, though some found ways to bracket or qualify its findings. Martin Delaney claimed there were still data he would like to see, but that it "wasn't worth fighting about." Concorde was "just one of many studies out there" and not a particularly relevant one, at that. Concorde had asked "one question": Does a daily gram of AZT initiated early in the course of disease prevent progression? But since no one routinely took a gram a day of the drug, it wasn;t clear "what that means for early intervention in general or even for AZT."
The big story, however, was the result from ACTG 155. This was Margaret Fischl's Phase II study that many had been awaiting in order to see whether the AZT/ddC combination, licensed the previous year, would perform as well in a long-term study as surrogate marker evidence had predicted. At a satellite symposium held in Berlin on the eve of the international conference, Fischl tried to put the best face on the study's outcome. The sad facts were that 42 percent of the subjects receiving AZT, 43 percent of the subjects receiving ddC, and 39 percent of the subjects receiving the combination had progressed to serious illness or death; there was no statistically significant difference between the three treatment arms of ACTG 155. The subjects receiving the combination therapy did get a larger CD4 count boost, just as they had in the promising early study, ACTG 106. But they didn't live any longer in the end.
Nevertheless, Fischl framed the outcome as positive by stressing that the combination therapy did show an advantage in the subset of patients whose CD4 counts at the outset were between 150 and 300 per cubic millimeter. Since these were the subject with the highest initial CD4 counts in the study, the "spin" that Fischl was pushing was that patients benefited from the combination unless they were already too far along in their illness. NIAID supported this interpretation, issuing a press release with the headline: "Effectiveness of AZT/ddC Combination Depends on Pretreatment Immune Cell Count."
Yet as the distressing word of the study's overall findings spread
through the halls of the conference center, reporters witnessed the unusual spectacle of a NIAID representative yanking the press release from the media center and hastily replacing it with a new, more sober one: "The Effectiveness of AZT Alone, ddC Alone or AZT/ddC Combination Is Similar Overall for Patients with Advanced HIV Disease." Fischl, undeterred, repeated her original conclusions at the formal conference session. No sooner had she finished than activists from TAG sprang to the microphone. "The answer to the study you designed is that the study shows no difference between combo and monotherapy," insisted a furious David Barr. "You have staked your career on these drugs," yelled Barr. "I have staked my life."
Barr and other activists characterized Fischl's stratification of the subjects by CD4 count as a methodologically unjustifiable violation of the fundamental principles of randomization, indeed, as a post hoc fishing expedition for results that would make the study appear a success. As much as activists had wanted to believe in the combination therapy, they now had no patience for any sugarcoating of the bad news. (Or perhaps, to the extent that they felt disillusioned, Fischl provided a convenient focal point for their anger: it's more satisfying, after all, to direct one's wrath at a researcher who appears to be cheating than at a virus that appears to be winning.) "How much is Roche paying you?" yelled out activists in the audience, referring to Hoffman-LaRoche, the manufacturer of ddC, which had already been the target of an activist protest during the conference's opening ceremonies because of the company's slowness in conducting its own, postmarketing studies of ddC as required by the FDA.
Whether Fischl was casting about to salvage her study or describing a genuinely important auxiliary finding was, in some sense, beside the point. Activist rage had more to do with the predicament the patient community now found itself in. As Mark Harrington noted in a heated challenge to Fischl voiced from the conference floor, the response to Concorde by U.S. experts had been to play it down, to claim that, after all, combination therapy was the real standard of care for asymptomatics. "The truth is," said Harrington bitterly, "we have no standard of care for asymptomatics." There was no disputing Harrington's point, nor did the conference presentations on non-nucleosides, tat inhibitors, and protease inhibitors suggest that answers lay just around the corner. Insiders already knew the gossip about Hoffman-LaRoche's tat drug: after years of stalling, the company had finally
determined that the drug had no efficacy, at least at the dosage tried so far. And the results from the protease inhibitors were preliminary and carried little weight.