Increased flexibility appears to be on the table as FDA turns to its Anti-Infective Drugs Advisory Committee for aid in re-crafting its guidances on developing antibiotics to treat pneumonia.

On Nov. 3 the panel will critique a two-step analysis process, with different non-inferiority margins at each step, for community-acquired bacterial pneumonia trials. The next day, the panel will assess whether a single trial, plus supporting data, is adequate to support an indication for hospital- and ventilator-acquired versions of the disease.

Practicality and feasibility issues raised in comments on earlier draft guidances for the disease prompted the look at alternative methods for evaluating antibiotics for the three indications, the agency explains in briefing materials for the meeting.

FDA issued the CABP guidance in 2009 (Also see "CAP Drugs Must Meet Stricter Noninferiority Criteria Under Draft FDA Guidance" - Pink Sheet, 8 June, 2009.), and the HAP document in late 2010 (Also see "Hospital-Acquired Pneumonia Is Life Or Death, FDA Says; Recommends Sole Primary Endpoint" - Pink Sheet, 7 December, 2010.).

One of industry's concerns about the CABP draft guidance is its focus on microbiologically confirmed cases and consideration of the microbiological intent-to-treat population as the primary analysis population (Also see "Former FDA Director: Pneumonia Standards May Be Tough To Follow In Real World R&D" - Pink Sheet, 15 September, 2009.).

Accumulating sufficient numbers in the mITT population could be difficult. FDA notes that prior discussions with AIDAC and other experts, comments on the draft guidances and the literature indicate only 25% to 30% of patients have a documented bacterial pathogen by conventional culture methods.

Pooling Of Trial Data For Final Efficacy Decision

To address this, FDA suggests sponsors could pool mITT data from two trials.

The two studies would need to have identical protocols. Efficacy of the antibiotic in the ITT population would be independently analyzed in each trial, using a non-inferiority margin of 10%. If each trial demonstrates non-inferiority, patients in the mITT populations in the two trials would be pooled and analyzed using an NI margin of 15% to show efficacy.

FDA's briefing materials also outline a development pathway that uses one NI trial in CABP, supported by a trial in HABP or acute bacterial skin and skin structure infections. The existing draft guidance suggests that sponsors could discuss submitting a trial in another respiratory infection as support for a CABP indication, but did not specify which ones.

CABP Endpoint Of Symptom Improvement

Also on the agenda are endpoints for CABP. While a better endpoint is developed, FDA is willing to look at improvement in at least two symptoms at days 3 to 5 after initiating treatment as a primary endpoint. Symptoms include cough, sputum production, chest pain and shortness of breath.

The agency considers this a possible interim endpoint while work is ongoing to characterize a clinically meaningful alternative. In particular, the agency encourages development of a patient-reported outcomes instrument to capture symptom-response, as well as improvement in clinical signs from the patient perspective.

In addition to the symptom-response endpoint, FDA says an essential secondary endpoint is improvement or stability in clinical signs, such as body temperature, blood pressure, pulse and respiratory rate. The agency also calls for an endpoint that reflects a patient's clinical status after completion of treatment as a secondary outcome.

The draft guidance called for success to be measured on the resolution of disease-specific signs and symptoms present at enrollment and no new symptoms or complications.

For HABP, One Trial Plus Supporting Evidence

FDA is willing to accept a single trial for drugs to treat HABP/VABP, plus supporting evidence, a change prompted by estimates of the sample sizes needed for two trials.

A single trial using an all-cause mortality endpoint could be supported by information such as evidence of efficacy and safety in other clinical conditions, Phase II trial(s) and data from efficacy in animal models of infections and in vitro studies.

A trial plan described in the briefing materials includes a single trial with supporting material from a trial in complicated intra-abdominal infections, CABP or acute bacterial skin and skin structure infections.

In addition to discussing the type of supportive evidence considered acceptable for a single trial, FDA is seeking panel input on whether an NI of 10% is acceptable if the active control mortality rate is <20% and="" if="" the="" odds="" ratio="" or="" risk="" difference="" metric="" is="" preferred.="" it="" is="" also="" asking="" about="" the="" timing="" for="" an="" all-cause="" mortality="">

The level of prior antibacterial treatment considered acceptable is up for discussion both days.

- Cathy Dombrowski (c.dombrowski@elsevier.com )

FDA Indicates Flexibility On Evidence To Support Efficacy In Pneumonia

More from Archive

More from Pink Sheet