Washington Editor

WASHINGTON - FDA drug reviewers Tuesday raised concerns about the design and analysis of Gilead Sciences Inc.'s efficacy studies intended to support approval of Cayston (aztreonam) as an inhalational antibacterial therapy for improving respiratory symptoms and pulmonary function in cystic fibrosis patients with Pseudomonas aeruginosa.

In briefing documents, which were released ahead of Thursday's meeting of the FDA's Anti-Infective Drugs Advisory Committee, regulators said they found problems with one study's patient-reported questionnaire and said the firm's re-analyses of data from another study could not be shown to be robust and would require making several problematic assumptions.

Aztreonam, a monobactam antibacterial, has been marketed in the U.S. since 1986 as an intravenous drug to treat a variety of infections due to Gram-negative organisms, including Pseudomonas aeruginosa.

Gilead in April 2003 submitted its application for Cayston, an inhalational form of the drug, based on two pivotal Phase III trials, CP-AI-005 and CP-AI-007, which used different designs to evaluate different endpoints of interest.

A third submitted study, CP-AI-006, was an open-label follow-on study designed primarily to evaluate the long-term safety of multiple treatment courses of Cayston.

Study CP-AI-007 was a double-blind, placebo-controlled trial of 28 days of a one-dose regimen of Cayston in cystic fibrosis patients 6 years or older with baseline forced expiratory volume in one second (FEV1) predicted between 25 percent and 75 percent.

The primary endpoint for the trial was change from baseline in the respiratory domain of the Cystic Fibrosis Questionnaire-Revised (CFQ-R) from baseline, or day 0 of drug treatment, to the end of study drug treatment, or day 28.

FDA reviewers noted that there were several shortened versions of the CFQ-R for children younger than 14, while the CFQ-R for adults and adolescents consisted of several demographic questions followed by 50 questions intended to measure health-related quality of life for patients in multiple domains.

Seven questions were related to the respiratory symptoms domain of the CFQ-R. But drug reviewers said there were several concerns about those questions and the responses, including missing data for patients who did not complete the CFQ-R.

Regulators also said there were several concerns about Gilead's analyses in study CP-AI-005, which generally related to the firm's post-hoc nature and the limitations involved with combining post-hoc evidence with confirmatory evidence.

After submitting its new drug application to the FDA, Gilead submitted a re-analyses to address the agency's concerns about the pronounced regimen effect observed in study CP-AI-005.

But the company's re-analyses of data could not be shown to be robust and would require making several problematic assumptions, FDA reviewers said.

"Since the applicant's re-analyses were conducted in a post-hoc manner and not based on pre-specified hypotheses, there are concerns that assumptions used in the analysis could have been determined retrospectively and primarily driven based on the significance of findings observed," regulators said.

If a large number of assumptions had been previously considered and rejected based on suboptimal findings of significance, serious inflation of the overall type I error rate was most likely, and therefore, the strength of evidence from post-hoc analyses was generally considered to be less reliable than evidence obtained from a well-controlled confirmatory study, the FDA said.

The FDA will ask its panel Thursday whether Gilead has provided substantial evidence of the efficacy and safety of 75 mg three times daily of Cayston as a therapy to improve respiratory symptoms and pulmonary function in cystic fibrosis patients with Pseudomonas aeruginosa or if further information would be needed before approving the drug.