ADRIA’s MYCOBUTIN (RIFABUTIN) RECOMMENDED FOR PREVENTION OF MYCOBACTERIUM AVIUM COMPLEX INFECTION; CLINICAL BENEFIT "MODEST" - - FDA ADVISORY COMMITTEE

Adria's Mycobutin (rifabutin) should be approved for the prevention of disseminated Mycobacterium avium complex (MAC) infection in HIV-infected persons, FDA's Antiviral Drug Products Advisory Committee recommended at its Sept. 24, meeting. The committee voted nine-to-zero, with one abstention, in favor of approving the drug for the prophylaxis of the bacterial opportunistic infection, which often occurs in the late stages of AIDS. The abstention came from Donald Abrams, MD, San Francisco General Hospital. The meeting was the second advisory committee review for the Adria drug. In June, the committee concluded that it had inadequate information to recommend rifabutin for approval and requested that Adria conduct further analyses of the clinical trial data ("The Pink Sheet" June 8, p. 3). During the approval vote at the Sept. 24 meeting, committee members emphasized that Mycobutin prophylaxis produced only a "modest" clinical benefit in trials. "I think that the drug should be approved and I think we have seen [an] indication of some clinical benefit this morning," committee member Deborah Cotton, MD, Harvard Medical School, said. The clinical benefit of the drug was "certainly nothing earth-shaking, but I think everything put together really suggests that this drug is preventing invasive disease," she added. Other committee members voted for approval but found the data marginally convincing. Committee member and statistician Paul Meier, PhD, University of Chicago, remarked: "I found the evidence much weaker than I think it ought to be, but, grudgingly, I guess I say I find some clinical benefit. Not a lot." Adria's trial design, which primarily tracked MAC bacteremia and survival, may have limited the trial's ability to demonstrate clear-cut efficacy, some committee members and consultants suggested. Noting that "the critical question was whether this surrogate marker [MAC bacteremia] was accompanied by the necessary clinical improvement," committee member Monto Ho, MD, University of Pittsburgh, said "the original study, particularly the follow- up, was not designed to answer this question." The effect of Mycobutin may be greater than that seen in trials, some committee members and consultants indicated, but had been obscured by the presence of various other opportunistic infections found in late-stage AIDS patients. Committee member John Modlin, MD, Dartmouth/Hitchcock Medical Center, remarked: "I think the benefit is somewhat larger than we can measure, although we don't know how much more so, but it is difficult to measure because of the background noise -- other opportunistic infections and primary infection [associated] with HIV disease." Adria presented further analyses of its 1,146-patient clinical trial database that the company said showed delays in developing MAC bacteremia and significant reductions in fever, fatigue and some laboratory abnormalities. The committee, however, based its approval recommendation almost entirely on FDA's analysis of company data. The committee agreed that FDA had "teased out" the clinical benefit of rifabutin by investigating the activity of the drug in relation to MAC bacteremia in a variety of symptoms. Without FDA's analysis, "one would have to be even less enthusiastic, if at all approving of the drug company analysis," Ho said. "FDA has performed an analysis for the company to put this drug on the market," he added. The agency analysis, presented by Antiviral Drug Products Division Medical Officer Robin Edison, MD, looked at MAC bacteremia alone and in combination with one or more symptoms thought to be associated with the infection. A panel of three experts consulted by the agency determined which symptoms were MAC-related. FDA also looked at overall survival differences between rifabutin and placebo-treated patients. Similar to Adria's analyses, Edison found that patients treated with rifabutin were half as likely to develop MAC bacteremia as patients taking placebo. In addition, Edison showed that rifabutin's prophylactic effect was statistically significant for all patients with MAC bacteremia and experiencing at least one symptom. The symptoms included: 5% weight loss; 10% weight loss; fever and night sweats; a hemoglobin decrease of 1 gram or more; increases in alkaline phosphatase; and elevation of liver transaminases. "Regardless of which sign you look at, the reduction in bacteremia plus that sign are all in the range of 1.5 to 2.5 [fold], with the exception of alkaline phosphatase, which is considerably higher," Edison reported. When FDA examined the effect of rifabutin on MAC bacteremia and more than one symptom, the agency found similar clinical improvements. Referring to MAC bacteremia plus two symptoms, Edison noted that "the reduction associated with rifabutin is in the range of two to four [fold], with the largest reduction seen in bacteremia associated with a decrease in hemoglobin and a greater than 10% weight loss." Concerning rifabutin's effect on survival, Edison observed that while reducing MAC infection should improve survival two-to- three fold, there was no significant effect of rifabutin on survival in the Adria trials. "Although it may be that there are some trends in [a] favorable direction, we did not feel that a survival benefit has been demonstrated," Edison noted. MAC "is only one of many factors reducing survival in AIDS, so preventing MAC may not have a major impact on overall mortality due to AIDS," she explained. Committee members differed on the question of at what point rifabutin therapy should begin in HIV-infected individuals. A majority of committee members felt that initiation of rifabutin prophylaxis should occur when a patient's CD4 count has decreased to 100 cells/ml or less since few patients develop MAC bacteremia above that CD4 level. Other committee members thought that therapy should begin at a CD4 count of 200 with an AIDS-defining event. The committee ultimately left the details about when to begin therapy to FDA. Agency medical reviewer Mark Goldberger, MD, suggested that the labeled indication for MAC bacteremia be followed by a description of the patient population in the clinical trials. Goldberger said the labeling could "describe who was studied in the trial and what we could say about the risk." Labeling could state that "patients with CD4 [counts] under 200 were included but, in fact, the great majority of the [MAC] cases appeared to occur [at CD4 counts] under 100," Goldberger noted. The committee recommended that Adria perform Phase IV studies following approval of rifabutin, including pediatric trials, drug interaction and further clinical efficacy studies. "I would hope that the company would be committed to continuing to study this drug and certainly do head-to-head trials with other drugs that come along," Cotton said. Adria plans to enter the clinic with a pediatric formulation of Mycobutin in October. The company has conducted pharmacokinetic studies to show that rifabutin's interaction with AZT and fluconazole is minimal. The firm is planning other drug interaction studies between rifabutin and methadone, clarithromycin, azithromycin and dapsone. FDA said it also would like to see interaction studies of rifabutin with oral contraceptives and clofazimine. There are few significant side effects associated with the drug. Safety data collected from patients enrolled in the company's Treatment IND program, which has accrued 2,065 patients since beginning in March, show "no serious or unexpected adverse events," the firm said.