Frequency refuses to hear the death knell
The company placed part of the blame for its mid-stage hearing loss failure on subjects exaggerating their level of hearing loss.
Frequency Therapeutics’ insistence on carrying on with development of FX-322 despite the hearing loss project having shown precisely no efficacy over placebo in a phase IIa study is doubtless spurred by the lack of any other clinical-stage asset in its pipeline.
Whether the decision will turn out to be a case of wisely keeping faith with an effective asset poorly served by an imperfectly designed trial or one of sheer desperation is another question. Shareholders have made their feelings known by hacking three quarters off the company’s value so far today.
“We have taken a meaningful step forward clinically, though not in the way we anticipated,” said David Luccino, the group’s chief executive, on a conference call today. Quite: interim data from the FX-322-202 trial showed FX-322, a small molecule designed to regenerate hair cells in the cochlea, to be ineffective – but Frequency believes the trial, not the project, to be at fault.
Back to phase I
The study enrolled 95 subjects with mild to moderately severe sensorineural hearing loss. All patients received four weekly intratympanic injections, with none, one, two or all four being FX-322 and the rest placebo. There were no discernible differences in hearing improvements between any of the groups.
Frequency theorised that the repeated injections were a problem, saying that these “seem to temporarily overwhelm the ear”, creating an unfavourable biological environment to improve hearing. It pointed to earlier trials in which a single dose of FX-322 had shown a significant benefit over placebo.
More startlingly, the group suggested that some of the responsibility for the trial’s failure fell on the participants themselves. There had been improvements in word recognition scores in placebo-treated patients and in the untreated ears of the active group patients – something that had not been seen in other trials. The company believes that some subjects had pretended that their word recognition was worse than it really was to get enrolled in the trial – hence the improvement in untreated patients.
Future trials of FX-322 will focus on single doses, the group said, and will be designed to eliminate enrolment bias. But for now the only upcoming readouts for the project come from phase I trials, one in age-related hearing loss and another in severe disease.
Even if these come up trumps the group will have to navigate phase II trials all over again. A lot might hinge on whether Astellas, which took ex-US rights to the project for $80m up front in 2019, opts to hand it back.
|Frequency's clinical trials to date|
|FX-322-202||Placebo-controlled, repeat dose phase IIa trial in mild to moderate SNL||95||Failed|
|FX-322-201||Placebo-controlled, single dose phase I/II trial in mild to moderate SNL||23||Significantly improved word recognition and words in noise vs placebo|
|FX-322-111||Open-label, single dose phase I safety study in mild to severe SNL||33||Significantly improved word recognition vs untreated ear|
|FX-322-112||Placebo-controlled, single dose phase I trial in age-related SNL||30||Expected Q2 2021|
|FX-322-113||Placebo-controlled, single dose phase Ib trial in severe SNL||30||Expected Q3 2021|
|SNL=sensorineural hearing loss. Source: EvaluatePharma, clincialtrials.gov, Otology & Neurotology.|