Design of trials for community-acquired pneumonia drugs would face greater specificity but no major changes according to recommendations from FDA's Anti-Infective Drugs Advisory Committee.

The panel agreed at an April 2 meeting that a non-inferiority margin of 10 percent in active-control non-inferiority trials should be acceptable to prove efficacy of CAP drugs.

FDA is working to update its guidance on CAP trial design and sought committee insight on how to translate the limited data from historic antibacterial trials into a scientifically verifiable treatment effect to serve as the basis for comparison to new antibiotics.

The committee discussed whether mortality data from CAP trials of penicillin, sulfonamides and tetracyclines in the early 1900s can serve to establish a treatment effect for comparator drugs on clinical outcomes and in different levels of disease severity.

Despite complications with data interpretation, the panel settled on the 10 percent non-inferiority margin after unanimously agreeing that a treatment effect can be quantified for CAP studies of IV drugs in hospitalized patients, and they voted 10 to 3 that the same could be said for an oral formulation in outpatient CAP patients.

Is Non-Inferiority Inferior In CAP As Well

The discussion of CAP clinical trial design is part of FDA's ongoing efforts to provide clearer guidance on antibiotic trial requirements.

Non-inferiority trials are a popular option for anti-infectives, but FDA has issued guidance requesting that superiority trials be conducted for several self-resolving community respiratory tract infections (¹ [A#00690430011], p. 17).

FDA expressed concerns about use of non-inferiority trials in CAP as well at a recent public workshop, noting that industry has provided the agency little data in support of the design (² Also see "FDA Still Needs Convincing On Use Of Non-Inferiority Trials For CAP – Temple" - Pink Sheet, 4 February, 2008., p. 30).

However, there is general consensus that conducting superiority or placebo-controlled trials in CAP patients is unrealistic. All clinical trials recently submitted to FDA for a CAP indication have been non-inferiority trials.

However, proper non-inferiority trial design in CAP is challenging. The treatment effect for active comparators is unknown because no placebo-controlled trials have been conducted for CAP since the 1940s; and the conduct of further placebo-controlled trials is widely considered unethical due to the known value of antibiotics. The panel voted unanimously against placebo-controlled trials even for mild to moderate CAP.

Moreover, active comparator superiority studies are unattractive to industry due to the likelihood of failure to demonstrate superiority in the self-resolving condition, even for effective drugs.

Thus, development of new agents to treat CAP falls to active comparator non-inferiority studies for which FDA is seeking to define an appropriate non-inferiority margin and patient population. The original CAP trial draft guidance was issued in 1998. FDA has no timeline for completion of the updated policy, but said it is a priority.

In the meantime, the fate of several pending or planned CAP submissions await, including those for Advanced Life Science's cethromycin, Wyeth's Tygacil (tigecycline), Johnson & Johnson's ceftobiprole, and Forest Laboratories' ceftaroline.

"One of the realities of our work is that we often have to give guidance to individual companies in the absence of having updated officially the guidance," FDA Office of New Drugs Director John Jenkins said following the meeting.

"If someone approaches us next week and says 'we want to now discuss with you our program for CAP,' we'll do that. We probably won't directly apply the 1998 guidance. We won't be able to directly apply the future guidance. But we'll have to give them our best advice today, and clearly what we heard from the committee informs what we will tell them today," he added.

What Is Equivalent To Death?

There was some disagreement among the panel about how to extrapolate new endpoints from the treatment effect determined by the mortality in the historical data.

All but one committee member felt that it is fine to rely on the prior clinical data to evaluate mortality for CAP candidate drugs.

"We have far more patients with underlying conditions and co-morbidities - it's very different right now from what it was in the 1930s. It's so difficult to interpret the data, I'm not sure we're doing anybody a favor by dealing with it," temporary voting member Daniel Musher, Baylor College of Medicine said.

The rest of the panel debated the appropriateness of evaluating measures of clinical efficacy for the new products.

Temporary voting member Thomas Fleming, University of Washington, argued that the extrapolation could not be statistically justified. "I think the fundamental principle that first needs to be recognized is that margins that you use are specific to endpoint. ... Justifying the margin using data for one endpoint to another is conceivably possible but enormously complicated."

He urged that "mortality is still a highly clinically relevant endpoint, as well as one that historical data provide us the best sense as to what the effect of the active comparator is, given that's the base to do a valid non-inferiority assessment."

"Sixty years ago they didn't have any rescue medications," Jürgen Venitz, Medical College of Virginia said, "so I do think there are other things that allow you to make that translation. I believe that we can translate quantitatively the mortality difference that they found 50-plus years ago to clinical success as we do it today."

FDA seemed open to translating mortality data to support trials evaluating clinical outcomes. "It seems very reasonable to take what constitutes failure in the modern world and equate that with mortality," Office of Drug Evaluation Director Robert Temple said.

In-Patients Of Yesterday Are Out-Patients Of Today

The panel agreed that a treatment effect can be quantified for non-inferiority studies of oral formulations for outpatient CAP. And there was general consensus that the same endpoints used for severe CAP could be used in testing for outpatient CAP drugs.

Those in favor of extrapolating the mortality data to the outpatient setting noted that modern medicine allows many patients that would have been hospitalized and placed on IV therapy at the time of the historical studies to be today treated as outpatients placed on oral antibiotics.

"I think everything we heard made us think that you continue to use that kind of endpoint here just as you would in the more severe illness," Temple said.

"What I heard is that people think that the data we have on severe illness is probably now applicable to people who aren't necessarily hospitalized because the modern world makes that unnecessary. But that really means they're fairly sick or have severe disease - we're not talking about things that you ordinarily think of as just really mild."

The panel also voted unanimously that establishing efficacy in severe CAP supports the drug's effect in less severe CAP, even though it has not been directly studied for that indication.

Using a Syndrome To Determine CAP Severity

The panel also weighed in on how to determine disease severity.

"The idea is to enrich for bacterial pneumonia," temporary voting member Scott Dowell, Centers for Disease Control, said. "The way you define CAP may be one of the most important things we do to enrich for those patients who actually have bacterial pneumonia."

"What seems reasonable and possible to do is you start out with a ... clinical syndrome that looks very much like a typical bacterial pneumonia," said AstraZeneca Infection Clinical VP John Rex, the non-voting industry representative.

"The classic syndrome needs to be an abrupt onset of febrile syndrome that points to the chest with a change on the chest x-ray, and probably with the ability to produce some sputum. Those things together suggest a bacterial etiology, and it's the acuity of the presentation," he said.

- Jamie Hammon (j.hammon@elsevier.com)

FDA Panel Offers Specifics For Revised CAP Trial Guidance, No Big Changes

More from Archive

More from Pink Sheet