The Questions of Real Importance
Of course, activists also had quite a bit of learning to do about clinical trials, and biostatisticians played an important role in that education process. Both Susan Ellenberg of NIAID and Mark Harrington of ACT UP/New York pointed to David Byar (who died of AIDS himself in August 1991) as the key "charismatic" figure who helped bridge the gap between worlds. Byar suffered no fools—whether statisticians, activists, or researchers—but was happy to expound on the inner logic of the controlled clinical trial until his audience grasped the issues. Inevitably, the greater exposure to the science of clinical trials caused activist opinions to begin to shift.
An interesting example concerned the debate over the ethics of placebos. The idea that placebo-controlled trials were inherently unjust in life-threatening illnesses had endowed treatment activists with a moral claim, and it had made for some catchy slogans. But ironically, the same argument that had helped mobilize the AIDS movement to pay attention to clinical trial design had little currency once activists sat down at the table with the wizards of statistics. Simply by accepting the terms of the debate—that properly conducted, controlled trials should guide the selection of therapies—treatment activists became bound, at least to some extent, by the inner logic of these evaluative methods. And that logic, as David Byar insisted, at times favored the use of placebos as the shortest route to a meaningful answer: placebos were "sometimes … in the patients' best interest, whether they realize it or not, and no matter how many signs they paint and march around New York City with. …"
As D. Bruce Burlington of the FDA explained, the FDA did not necessarily insist on placebo controls as the only mechanism for conducting a controlled study that could lead to a drug's approval. For
example, "the agency can, and frequently does, accept … the historically controlled trial," where patients receiving an experimental treatment are compared with a matched group of untreated patients who were followed at some point in the past. But drug sponsors relying on historical controls "bear the burden of establishing the natural history of the disease," that is, proving that the progression of illness within the untreated cohort was representative of the disease in question. The problem, according to Burlington, was that AIDS had no single, constant natural history, given "the dramatic changes in demographics and the marked improvement in diagnosis and management of [HIV-infected] patients." The natural history of AIDS was in flux; so what would the treatment group be compared against?
Some activists, like Jim Eigo, rethought their assumptions: at another panel discussion in 1991, he acknowledged that although he originally had seen no need for placebos ever, he now recognized the virtues of using them in certain situations, where a short trial could rapidly answer an important question. More generally, activists maintained that the point was not so much placebos versus no placebos, but whether a trial asked a meaningful, real-world question that the patient community wanted answered. As Eigo put it in 1989 at a symposium entitled "Methodological Issues in AIDS Clinical Trials," sponsored by NIAID and the FDA: "If every arm of every trial asked a question of real importance to people with acquired immune suppression, enough of those people would find every arm of every trial a viable treatment option and, therefore, if they knew about the trial, could be accrued for that trial." This, in a sense, was a new definition of "equipoise" reframed from the vantage point of the community of patients rather than the community of scientists. If a trial compared different treatment options (including one option that was the control), if patients didn't know which option was best, if patients wanted to know which option was best, and if every option provided patients with quality medical care in all other respects, then patients would sign up for the study, regardless of the type of controls used. The prescription, therefore, was for better and more profound communication between researchers and the subject population (or their activist representatives) before studies got off the ground—indeed, before they were even proposed. "When we talk about methodology," Mark Harrington told the participants at the "Methodological Issues" conference, "we usually talk about how we are going to answer the questions that we set out to ask." This begs the prior question of "how
… we decide what are the important questions. As patients become more involved in the design and execution of clinical trials," said Harrington, "it is crucial to recognize that patients' questions are very important and deserve to be answered."
The questions that regulators or researchers wanted answered were often too academic, too removed from the day-to-day realities of patient care; and the resulting trials didn't provide participants with attractive options. Only an active engagement and negotiation between researchers, doctors, and the AIDS movement could ensure that the most important therapeutic questions were being studied; only such an engagement could rescue researchers from the not infrequent difficulties in recruiting patients to participate in their trials. As Paul Volberding acknowledged in his presentation at the same conference, "examples abound of clinical trials that are elegant in their design but fail because of limited accrual of subjects."