Two pivotal trials of Merck’s chronic cough project gefapixant were top-lined as successes in March. Now the full data are out, and the P2X3 receptor inhibitor’s efficacy does not look quite so impressive.
The drug reduced coughing, but to a much less dramatic extent than had been seen in phase II. It is probably too soon to write off gefapixant’s chances of approval, but these data, coupled with the drug’s poor side-effect profile, could limit the potential market.
According to the full data from the Cough-1 and -2 trials, presented at the virtual congress of the European Respiratory Society, the higher gefapixant dose of 45mg twice daily demonstrated a statistically significant – just – reduction in 24-hour cough frequency versus placebo. But, at 18% and 15% in Cough-1 and -2 respectively, these reductions, calculated using a complex formula, are disappointing compared with what had been achieved in earlier trials.
|Phase III gefapixent data|
|Cough-1 (week 12)||Cough-2 (week 24)|
|Placebo||Gefapixant 15mg||Gefapixant 45mg||Placebo||Gefapixant 15mg||Gefapixant 45mg|
|Baseline 24hr cough frequency (coughs/hr)||23||20||18||19||19||19|
|24hr cough frequency at primary timepoint||10||10||7||8||8||7|
|Estimated relative reduction (%) vs placebo*||-||2||-18||-||-1||-15|
|*Estimated by 100*(exp(diff) -1), where diff was the difference provided by the analysis of the log transformed variable. Source: adapted from Merck press release.|
In phase II a 50mg dose of gefapixant achieved a 38% placebo-adjusted reduction in 24-hour cough frequency at 12 weeks. It should be noted that the relative reduction calculation for this figure was not made explicit.
Industry watchers will note that the cough reductions with gefapixant look pretty similar to those achieved in the failed phase II trial of Bellus’s rival cough therapy, BLU-5937. A 200mg twice-daily dose reduced awake cough frequency – slightly different measure to the 24-hour cough frequency endpoint used by Merck – by 17% (Bellus chokes in chronic cough, July 06, 2020). Today Bellus outlined its plans for a phase IIb trial of ’5937, focusing on particularly badly affected patients.
The phase III data on gefapixant are also underwhelming when patients’ improvements over the course of the trial are considered. The mean reduction from baseline in cough frequency with the 45mg dose was 62% in Cough-1 and 63% in Cough-2. However, placebo-treated patients saw reductions from baseline of 55% and 57% in Cough-1 and -2, respectively.
Also concerning is the fact that efficacy seems to be waning over time. Patients in Cough-2 received the drug for six months, twice as long as those in Cough-1, but achieved a smaller effect.
Then there is the tolerability aspect. Gefapixant is notorious for its unpleasant side effects of alteration or loss of taste, and in Cough-1 and -2 respectively 58% and 69% of patients experienced taste-related adverse events. 15% and 20% of subjects, respectively, discontinued therapy because of adverse events.
The data will be filed with regulators, Merck said, and such is the unmet need in refractory or unexplained chronic cough that approval might be possible – the higher dose did hit the study’s endpoint.
The difficulty could come in convincing doctors that the mild treatment effect is worth the side-effects. Gefapixant is currently forecast to sell $165m in 2026, according to sellside consensus from EvaluatePharma. This figure might now be trimmed further – perhaps to the point that Merck’s investors start wondering why the company is bothering with it.