VALIDATION OF CD4 COUNT IN AIDS DRUG DEVELOPMENT WILL BE DISCUSSED BY ANTIVIRAL ADVISORY CMTE. IN JUNE; CMTE. REMAINS WARY OF RELYING ON CD4s

FDA's Antiviral Drugs Advisory Committee will discuss the use of CD4 cell counts in the development of drugs to treat AIDS during a June meeting, Center for Drug Evaluation and Research Director Carl Peck, MD, said during the committee's April 20-21 meeting. An exact date for the June meeting has yet to be set. FDAers mentioned the future meeting during a brief update on the agency's work to validate the use of CD4 cell counts as a surrogate marker. "In June, we are planning a more lengthy time to discuss this," Peck said. FDA's presentation on CD4 counts followed the committee's determination that clinical data demonstrated the anticipated benefit of Bristol-Myers Squibb's Videx (ddI), which was approved based on surrogate marker data. The presentation preceded the committee's review of Hoffmann-La Roche's application for Hivid (ddC) (see related stories, p. 14-19). Following the update, Peck told the committee that they "should come away from [the April 20-21] discussion with a fairly positive view about the prospect of a linkage between CD4 and outcome." FDA staff presented models for validating CD4s as an endpoint and applied those models to data from the Burroughs Wellcome 02 study which demonstrated a survival benefit for AZT. CD4 cell counts correctly predicted outcomes in 77% of cases, Peck summarized. Moreover, he said, the FDA analysis is the first to "rigorously demonstrate" that "change in the CD4 enters independently as a variable for outcome." The success of the models in analyzing the BW 02 data is a "lower bound" of their strength, Peck added, since more recent trials are likely to have sounder CD4 measurements. At the June meeting, investigators involved with the National Institutes of Health's AIDS Clinical Trials Group studies of ddI plan to present correlation analyses of changes in CD4 counts and their relationship to clinical outcomes. The investigators had hoped to provide the analyses at the April 20 meeting, but they ran into complications in analyzing the data. Another contribution to the topic for discussion may be the report of FDA's task force on surrogate endpoints, which is expected to be completed in late May. Theodore Eickhoff, MD, Presbyterian-St. Luke's Medical Center, Denver, characterized the committee's discussion April 20 as a "litany of discomfort" that recognized the panel's continued ambivalence about the use of CD4 cell counts. The committee generally supported the use of CD4s as "partial but significant" surrogate markers at its February 1991 meeting ("The Pink Sheet" Feb. 18, 1991, p. 8). The committee then relied on CD4 data to recommend approval for ddI pending the outcome of clinical efficacy trials ("The Pink Sheet" July 22, 1991, p. 3). However, at a joint meeting with the Vaccines & Related Biologicals Committee in November, the use of CD4 counts as a primary endpoint for non-nucleosides was rejected ("The Pink Sheet" Nov. 18, 1991, p. 17). Discussion April 20 focused on perceived drawbacks of the use of CD4. Donald Abrams, MD, San Francisco General Hospital, said: "I continue to be concerned about the lack of real translation from these CD4 changes to survival." Mark Smith, MD, VP-Henry J. Kaiser Family Fund, said he was concerned about "which functions of CD4" are most predictive: absolute change, percent change, area under the curve, etc. He later characterized the presentation of CD4 data from the ddC application as "the beginning of some selective use of CD4s when it seems to show an effect and ignoring them when they don't." A majority of the committee, however, appeared to agree with the position expressed by Fred Valentine, MD, New York University Medical Center, who said: "At present, I think drug-induced CD4 changes for nucleoside analogs are an appropriate surrogate to give an early approval. Subsequent approval and really convincing evidence as to the efficacy of the compound would have to depend on clinical endpoints." Chairman Henry Masur, MD, NIH Clinical Center, concluded, "We didn't have a consensus on ddI, and we don't have one now."