US FDA Makes Flexibility Look Routine In TB Drug Review, But Not Everyone Is Pleased

The most remarkable aspect of the positive advisory committee for pretomanid is how routine FDA’s flexibility in the face of unmet need is starting to feel – but there are some dissenting voices that merit attention. 

The US FDA is making regulatory flexibility look like a snap.

The most remarkable aspect of the US Food & Drug Administration’s approach to the review of pretomanid, a potential new treatment for highly resistant strains of tuberculosis submitted by the TB Alliance, is how routine it seemed.

Pretomanid has a Qualified Infectious Disease Product designation, signaling its special status in the context of the campaign to address microbial resistance. Like so many recent FDA approvals, the drug is focused on a significant unmet need in a small patient population, at least in the US where multi-drug resistant TB is extremely rare. Pretomanid would provide another backstop to treat cases that do turn up domestically, but FDA approval is intended to allow for availability in global regions where the infections are more common.

Now, after a positive (14-4) advisory committee vote on 6 June, it appears on track to become the first new TB drug approved in the US since J&J’s bedaquiline in 2012. The committee vote was very much in line with FDA’s positive review released ahead of the meeting, and agency officials did nothing to complicate the discussion and vote on June 6.  (Also see "Pretomanid Gets Panel Endorsement As Part Of Regimen, Not Individual Molecule, For Treatment-Resistant Tuberculosis" - Pink Sheet, 6 Jun, 2019.)

The straightforward treatment of the application in the FDA review is itself remarkable as an indication of the agency’s comfort level with clinical development plans that stray far afield from the mythological “gold standard” of two large, placebo-controlled trials.

The FDA's straightforward treatment of the application is itself a remarkable indication of the agency’s comfort level with clinical development plans that stray far afield from the mythological “gold standard” of two large, placebo-controlled trials.

Pretomanid has a non-traditional sponsor and relies on an unusual clinical trial analysis plan. The Global Alliance for TB Drug Development is a non-profit drug development company funded by the Gates Foundation and other international entities; it has entered into a non-exclusive agreement with Mylan NV to manufacture and commercialize the therapy.

 

The pretomanid NDA relies on one, single-arm trial that was halted after about half the planned patient population was enrolled, with the efficacy data derived from the first half of that smaller patient population, and supplemented after the NDA was filed.

That said, the reported results were impressive. The primary efficacy endpoint was bacteriologic failure, relapse, or clinical failure through the treatment and six-month follow up period. In the initial data set, 40 out of 45 patients (89%) had favorable outcomes. In the updated submission, 72 out of 80 evaluable patients did (90%). Both analyses showed favorable outcomes well above the pre-specified threshold of 50%, even when FDA conducted analyses to consider for screening failures and multiple endpoint testing penalties.

Still, it is very easy to imagine a different regulatory approach that focused on the flaws rather than the (apparent) benefits.

First, the 50% threshold was justified by a literature review and a matched case-control study conducted by the sponsor. But TB treatment is not static, and two of the open public hearing speakers June 6 argued that the background rate of cure is likely much higher in current practice.

Then there is the fact that the regimen studied by the TB Alliance combines pretomanid with the relatively recently approved bedaquiline – and the antibiotic linezolid, which is not approved for TB in the US (though it is included in WHO treatment guidelines).

A notable voice of caution came during the open public hearing from a speaker who is among the most influential patient advocates in FDA history, Treatment Action Group founder Mark Harrington.

FDA’s reviewers did not let those issues overly complicate the issue of what appears to be a highly effective and relatively convenient treatment in a setting of dire medical need. But there were other voices of dissent during the June 6 meeting from some perhaps surprising sources.

 

One of those voices of caution came during the open public hearing from a speaker who is among the most influential patient advocates in FDA history, Treatment Action Group founder Mark Harrington.

Harrington drew on his experience from HIV advocacy to offer an analogy. No one would have approved protease inhibitors to treat HIV if the studies were based on historical comparisons to the 1980s before any antiviral regimens had been developed, he argued. That, he suggested, is similar to the case for approving pretomanid based on assumptions about extremely low cure rates from before bedaquline and linezolid were used in resistant disease.

Patients deserve sound scientific evidence of benefit for new therapies, Harrington declared – and the level of unmet need and vulnerability of patients only increases their right to strong science. Perhaps fortunately for the TB Alliance, Harrington did not explicitly urge rejection of the application.

The most important voice of dissent, however, came from one of the four “no” voters on the committee: University of Pennsylvania biostatistics professor Susan Ellenberg. 

University of Pennsylvania biostatistics professor Susan Ellenberg seemed genuinely shocked by the FDA review team’s matter-of-fact handling of the application.

Ellenberg is one of a handful of go-to statisticians on FDA advisory committees whose academic work on clinical studies helped establish the framework for efficacy analysis. She is usually a strong but measured voice on committees – but delivered an unusually passionate “no” on pretomanid.

 

She explained her vote by re-reading the voting question: “Has the applicant provided substantial evidence of the effectiveness and sufficient evidence of the safety of pretomanid as part of a combination regimen with bedaquiline and linezolid, in adults for the treatment of pulmonary extensively drug resistant (XDR) or treatment-intolerant or nonresponsive multidrug-resistant (MDR) tuberculosis?”

“Substantial evidence?” she repeated. “I have no idea what pretomanid contributes to this regimen.” She noted that the only direct comparison of the three drug regimen to the component parts was conducted in mice. She pointed out that if murine models are sufficient to prove efficacy, then most drugs that enter clinical trials would end up approved.

She seemed genuinely shocked by the FDA review team’s matter-of-fact handling of the application. “I’m very disappointed that there was an agreement to consider this drug,” she said. “The desperate situation” for resistant TB patients “is no excuse,” she said. “People deserve drugs that work.”

Those voices of dissent were by far the minority during the course of the 6 June review. But they may suggest that no one should count on FDA making extreme regulatory flexibility seem so routine forever.