FUJISAWA AMBISOME COULD BE FIRST EMPIRICAL THERAPY FOR PRESUMED FUNGAL INFECTION; FUTURE TRIALS SHOULD DEFINE PRESUMED INFECTIONS, FDA COMMITTEE SAYS

Fujisawa USA's Ambisome could be the first approved empirical therapy in febrile neutropenic patients, following a unanimous vote by FDA's Antiviral Drugs Advisory committee July 16 that the liposomal amphotericin B product is safe and effective for the indication. Fujisawa USA licenses the product from Boulder, Colo.-based NeXstar Pharmaceuticals.

Although other amphotericin B products on the market are used off-label as empirical therapy for fungal infection in patients with febrile neutropenia, Ambisome could be the first to gain the labeled indication. Sequus' Amphotec liposomal amphotericin B was not recommended for the indication by the antivirals advisory committee April 14 because the company did not have adequate data to demonstrate equivalent efficacy to the standard amphotericin B treatment ("The Pink Sheet" April 21, T&G-6).

The Liposome Company estimates that 40% of prescriptions for its liposomal amphotericin B product Abelcet are for empirical therapy even though the product is not indicated for that use. "A label claim for empiric therapy would likely be influencing that use of all lipid-based amphotericin B products," TLC said at a Montgomery Securities meeting in New York July 16.

The advisory committee meeting marked the first time the group met to discuss a drug under review by the Office of Drug Evaluation IV's new Division of Special Pathogens and Immunologic Drug Products ("The Pink Sheet" May 26, T&G-1). It will also be the last antivirals committee meeting presided over by David Feigal, MD, as ODE IV director; Feigal is moving to the Center for Biologics Evaluation & Research as medical deputy director .

Although the committee discussed only the empirical therapy for presumed fungal infection at its meeting, proposed indications in the Ambisome NDA also include treatment of systemic and/or deep mycoses including aspergillus and candida; treatment of fungal infections refractory to or intolerant of traditional amphotericin B; use in patients with renal insufficiency; prophylaxis against systemic fungal infections following chemotherapy or liver transplant; and treatment of visceral leishmaniasis. Fujisawa has ongoing studies of the product for histoplasmosis in AIDS patients and for cryptococcal meningitis in AIDS patients.

FDA Immunologic Division Acting Director Mark Goldberger, MD, noted that "the division has virtually completed its review of the data supplied by the applicant for antifungal therapy and has concluded...that the product has demonstrated activity." He added that "it's really not possible at present to define the level of this activity vis a vis the comparator" amphotericin B, "so we will not be making a statement about that."

Fujisawa submitted the Ambisome NDA on Nov. 8, 1996; it was filed by FDA Jan. 11. Results of a single pivotal trial of the product for empirical therapy were submitted as an amendment to the NDA on April 25. The NDA also contains data demonstrating antifungal activity from two European open-label studies. The division has "not completed any types of labeling discussions with the company," Goldberger said.

The 687-patient pivotal study compared a standard dose of Ambisome, 3 mg/kg/day, to a standard dose of amphotericin B, .6 mg/kg/day, although Ambisome doses ranged from 1.5 mg to 6 mg and amphotericin ranged from .3 to 1.2. The overall success rate was 50% for Ambisome (n=171) and 49% (n=169) for amphotericin B, principal investigator Thomas Walsh, MD, National Cancer Institute, told the advisory committee. Success was based on a composite primary endpoint of survival at seven days post-treatment (93% Ambisome v. 90% amphotericin B); resolution of fever resolved while neutropenic (58% for both groups); no new fungal infection (proven or presumed) (86% for both groups); no discontinuation due to toxicity or lack of efficacy (86% Ambisome v. 81% amphotericin B); and resolution of baseline fungal infection (9/11 for Ambisome, 8/11 for amphotericin B).

Safety data showed Ambisome to be less nephrotoxic than amphotericin B, with creatinine levels raised >= 1.5 times baseline in 29.4% of Ambisome patients compared with 49.4% of amphotericin B patients, and >= 2 times baseline in 18.7% of Ambisome patients v. 33.7% of amphotericin B patients. Hypokalemia was experienced by 6.7% of Ambisome patients and 11.6% of amphotericin B patients. Hepatotoxicity was similar in both groups, with 17.8% for Ambisome and 20.3% for amphotericin B.

Infusion related reactions were also lower in the Ambisome group. Fever spike for the first infusion was significantly higher for the amphotericin group -- an increase of >= 1C was observed in 16.9% of the Ambisome group compared with 43.6% of the amphotericin B group -- as was the incidence of chills (18.4% v. 54.4%). Vasodilation was seen more often in the Ambisome group than the amphotericin group for all infusions (5.2% v. .6%).

Overall, 9.6% of the Ambisome group discontinued due to toxicity, compared to 13.7% of the amphotericin B group. For both groups, 7.3% of all patients discontinued for non-infusion adverse events; Ambisome patients withdrew at a lower rate for infusion related reactions (1.5% v. 6.1%).

Study data showed that there were 11 proven emergent fungal infections for patients on Ambisome and 27 for amphotericin B. However, presumed fungal infections emerged in 33 of the Ambisome group, but in only 14 of the amphotericin B group.

"The idea that one could actually see an important result in proven fungal endpoints is something that is welcome...but not entirely expected," Goldberger said. "Even less expected was the finding that...the proven endpoints are in one direction, the presumed in the other....Perhaps we were not as entirely prepared prospectively as we might have been in terms of looking at this issue," he said.

Several committee members, including committee Chairman Scott Hammer, MD, Harvard Medical School, responded that the fact that the mortality rate for Ambisome-treated patients with presumed fungal infections was not as high as those receiving amphotericin B was "reassuring": 15% of Ambisome patients with presumed emergent fungal infections died, compared to 29% in the amphotericin B group. The death rate among patients with proven fungal infections was 33% for Ambisome and 46% for amphotericin B.

"There are lots of things we don't know about those patients that may have put them into a presumed category," Pamela Diaz, MD, Chicago Department of Health, said. "So it may be just an incidental result that they happened to end up in the Ambisome group."

Other committee members thought Ambisome's effect may have interfered with a diagnosis. Committee member James Lipsky, MD, Mayo Clinic, felt that if "there was some obscurity because of the therapy, it would most likely be because the therapy was better than the gold standard." Committee members agreed with Lipsky that "if one is going to make assumptions about presumed infection then there should be absolutely rigorous criteria for what that would be."

Overall, Hammer said, "this was a rigorously done study that was attempted to be designed with the best advice at the time and designed as an equivalence study and I think achieved its goals." He said "my own appreciation of the data is that for adults Ambisome is safer with respect to particularly nephrotoxicity and infusion related reactions."

Efficacy "equivalence is shown fairly clearly to the gold standard amphotericin B," Hammer continued. "There may be some superiority when you get to the proven fungal infections, a very important aspect. Although these numbers are small as a subgroup, they are a very important part of this study," he added.

The committee revisited the issue of future trial design, which was discussed at the April 14 meeting. "I don't think you can get away from a combined endpoint in these studies, in part for the practical issues of sample sizes, but also that that reflects the type of clinical situation in which these drugs are used," Hammer said. The trials could be "refined," Hammer noted, by focusing more on proven emergent fungal infections and survival. "One way to tackle that is to enrich the population with the higher risk patients so that you in fact try to ensure that you are able to see sufficient numbers of proven fungal infections to see a difference," he said.