Executive Summary

Large, simple clinical trials in lieu of expanded access programs would ensure both safety and efficacy data and reduce costs, National Institute of Allergy and Infectious Diseases AIDS Division Chief of Biostatistics Susan Ellenberg told a Nov. 30-Dec. 1 World AIDS Day conference on expanded access in Cambridge, Mass. "I think that we should be thinking about providing access to experimental drugs when at all possible within the context of a simple clinical trial," Ellenberg said. The conference was sponsored by the American Society of Law & Medicine, the Center for the Study of Drug Development, and the Harvard AIDS Institute. Commenting on the debate over whether the goal of expanded access programs should be treatment or research, Ellenberg said "it seems to be a great waste to write off the possibility of accomplishing both at the same time." She added: "It is essential to collect some minimal data on adverse events and safety and toxicity when relatively few people have received these drugs." Ellenberg cited the use of very large, centrally administered clinical trials in Europe to support the idea for such an approach in AIDS. "The analogies to these trials are limited in regard to the situation we now have in AIDS," Ellenberg acknowledged, "but there have been such successful trials conducted and I think it would be worth a lot of time and effort trying to see how we could do them." "Such trials would take relatively little extra effort than the kinds of expanded access programs that we have now, and I certainly think they could be accomplished with less in the way of data collection than [the ddI (Merck's Videx) pilot expanded access program and Treatment IND programs]," Ellenberg declared. She noted that expanded access programs "are very expensive and don't bring us any closer any faster to finding out if a drug works." Encouraging large clinical trials, Ellenberg argued, would also provide some incentives to industry that may be lacking in the expanded access concept. Large studies, she said, "would provide an additional major opportunity to study efficacy in a context where the results would be maximally generalizeable." The benefits to industry sponsors, Ellenberg suggested, would "essentially be a Phase IV study at a Phase II time." In addition, such studies could speed approval of drugs by providing enough efficacy data on which to base approval for "fast-track" AIDS drugs, she asserted. FDA Office of Drug Evaluation I Director Robert Temple has also expressed his support for such an approach. Temple first offered the idea at an Institute of Medicine roundtable meeting in March. During the summer, Temple told an industry group that "there is growing interest in expanding access through use of low-tech, or large, simple clinical trials" ("The Pink Sheet" June 11, p. 10). Ellenberg envisions the design of such studies to include minimal entry criteria and minimal limitations on the use by patients of concomitant medications. The studies would be centrally administered and, she suggested, easy to randomize by dosage levels. However, she also acknowledged that there would be some problems: "Data collection does add to the burden of treating physicians [and] data quality and reliability would be an issue," she remarked. Ellenberg was one of a group of statisticians involved in NIAID's AIDS Clinical Trials Group that recently published a report, "Design Considerations for AIDS Trials," in the Nov. 8 New England Journal of Medicine. Among other things, the group suggested situations in which uncontrolled Phase III AIDS drug trials would be justified and noted that the three phases of drug evaluation do not have to be done consecutively.