Unclear if Chronic Cough Drug Has Meaningful Benefit, FDA Reviewers Say

It remains “unclear” whether the modest effects of gefapixant, an investigational drug for chronic cough, represent a clinically meaningful benefit, said FDA staff ahead of a meeting of the agency’s Pulmonary-Allergy Drugs Advisory Committee.

On Friday, committee members will wrestle with exactly what constitutes a meaningful benefit for this incredibly common condition, as no threshold has been established, and whether the P2X3 receptor antagonist’s impact would even be perceptible to patients.

Last year, FDA rejected gefapixant for the proposed indication of treating refractory or unexplained chronic cough in adults. Beyond concerns about the drug’s modest effect, the agency took issue with the measurement system used to track patients’ coughs in the two pivotal trials and requested additional data — essentially asking that sponsor Merck recount coughing frequency with a different methodology.

In the new recount analyses requested by FDA, gefapixant at a 45-mg dose showed a 15-17% relative reduction in 24-hour cough frequency over placebo at week 24 (the two phase III trials’ primary endpoint, though that difference was significant in only one of them).

Subsequent analyses to better understand the treatment effect failed to move the needle, according to the agency reviewers. “Assessed a variety of ways, the reduction in cough frequency was consistently small,” they wrote in their briefing documents for the meeting.

Other concerns raised in the pre-meeting package included interpretation of supporting patient-reported outcome (PRO) data, and the potential unblinding of patients as a result of taste disturbance. That common side effect of gefapixant occurred in 65% of study participants and was a cause of early treatment discontinuation in 14%.

No therapies are currently approved to treat chronic cough, which affects an estimated 5% to 10% of all adults. It is considered refractory when not relieved by treatment for an existing condition that causes coughing — such as chronic obstructive pulmonary disease (COPD) or asthma — or simply unexplained when no underlying medical condition is present.

A host of products are used off-label for the condition (neuroleptics, opioids, local anesthetics), but they can carry risks and evidence supporting their use is limited. “As such, FDA anticipates that a new product approved for [chronic cough] will be widely used given the prevalence of the condition and lack of therapeutic options,” the agency reviewers wrote, noting that long-term treatment would be anticipated.

While the exact cause of chronic cough is unclear, “a variety of nociceptors (including purinergic receptors such as P2X3) and mechanoreceptors have been implicated as ‘cough receptors’ in the respiratory mucosa, which respond to both intrinsic and extrinsic noxious stimuli, as well as mechanical stimulation,” wrote agency staff. “As an antagonist of the P2X3 receptor, which is expressed on sensory neurons in the afferent limb of the cough reflex, gefapixant is hypothesized to ameliorate this increased sensitivity to noxious stimuli, which could suppress the cough reflex.”

Beyond the issues with the primary endpoint in the pivotal trials, secondary endpoints did not provide additional support for gefapixant’s effect on coughing frequency, according to the document, though some PRO data did suggest patient improvement.

In one of the trials, a significantly higher proportion of patients had a 1.3-point or greater increase on the Leicester Cough Questionnaire total score from baseline to 24 weeks (OR 1.4, 95% CI 1.0-2.0). But this represented just 3.3% more patients versus placebo, a numerically small difference “of questionable clinical significance,” noted agency staff.

At Friday’s meeting, panelists will weigh in on these issues and vote on whether the supporting evidence demonstrates that gefapixant provides a clinically meaningful benefit for this patient population. While the FDA typically follows the advice of its advisory committees, it is not required to do so.