Frequency Therapeutics — Hearing Loss Regeneration

Statistical significance is twice the bar of clinical significance. Clinical significance is considered a 5 word improvement, statistical significance improvement is considered 10+ words.

 

Thank you for correcting me. The results will indeed be published in the second half of next year, not the first.

 

TL;DR: 2023 is do or die for Frequency Therapeutics. There is something here but the results haven't been as consistent as desired and there are a ton of questions remaining. They only have $111M left in cash and operating expense were $98M last year.

If it doesn't see strong enough results to move towards market or at the least attract further capital it will sit on the shelf for quite while.

 

Even if their hearing drug flops, their MS drug might save them; who knows.

 

Don't lose the forest for the trees though. They have seen consistent results from their Phase 1/2 study & their open label study where they saw up to a 40% response rate in particular etiologies & severities. It only feels inconsistent because their later probing studies targeted etiologies & severities that weren't included in the earlier studies and it so happens that these populations didn't show a robust response. The chart I have attached breaks down the response rates they've seen by etiology & severity. The orange dotted circle indicates the populations they are targeting for the current 208 trial.

Funding wise, FREQ has slowed their cash burn rate by laying off 30% of staff as well as subleasing out half of their building to save a few million a year so the company has done a good job trying to preserve capital. They also have $90 million coming to them once Astellas initiates a Phase 2b overseas and another $140 million coming to them when they start Phase 3 so depending on the timing of those payments, they may not need to raise additional funding before FX-322 makes it to market.

 

I gotta be careful here. I don't want to be pouring cold water on the efforts of free enterprise who are trying to bring a workable product to the market. They are a welcome addition to the halls of academia who come out with these reams and reams of learned papers but as for curing hearing loss or tinnitus, they don't seem to be able to hack it.

But to sum up much of the discussion... and I'm sure this has been said before:

Concentrating on early onset tinnitus... well, a generous percentage of this group are going to be completely cured or they reach habituation anyway in the course of 12 months till chronic status sets in.

And

If FX-322 tackles the higher frequencies' hair cells, it should be able to cure or alleviate the high frequency tinnitus patients, no?

It could very well be that they are on the right track, but that there are some other co-factors involved.

Roll on the next trials. We gotta stay optimistic.

 

Too many people have too high of expectations for this space. Expecting them to cure hearing loss with the first ever drug to be potentially approved in this space is like expecting the field of oncology to have cured cancer on their first try.

FREQ was the first one that was able to achieve a >2000 fold expansion of cochlear hair cells in the lab. The failure to successfully expand cochlear hair cells in the past was the largest bottleneck that was stifling innovation in the hearing space as it was extremely difficult to get hands on enough cochlear hair cells to be able to successfully test compounds against. Due to FREQ's contribution to the space, it is now much easier for researchers and companies to grow cochlear hair cells to test compounds against.

Here is the study outlining the process they used:

Clonal Expansion of Lgr5-Positive Cells from Mammalian Cochlea and High-Purity Generation of Sensory Hair Cells

So it is not that they contributed a paper or two to the field of hearing loss research, they laid the groundwork that many hearing loss focused biotech companies of today are built upon.

 

Tony Ricci, Vestibular nerve caps, type 1 and 2 hair cells. Compound Action Potential (CAP).

 

@Chad Lawton, what is it you see in previous FX-322 trials that go above and beyond the results of the OTO-413 trials?

From memory OTO-413 had 57% responder rates on the safety trial. 40% on the efficacy trial. No failed OTO-413 trials or excuses for failed trials i.e. too many doses, participants lied.

All I see are weak but maybe positive signals for FX-322.

 

I don't really compare OTO-413 to FX-322 as it is comparing apples to oranges in my opinion. They work on completely different parts of the underlying biology. It also depends on which hat you want me to wear.

If I put on my personal hearing loss/tinnitus sufferer hat, OTO-413 sounds promising but I don't think I will be able to benefit from it. Same goes for FX-322, I doubt I will benefit from it.

But if you ask me to put on my investor hat which I wear most of the time since I'm heavily invested in FREQ, I think FREQ will be the winner versus OTIC in the long run based on the fact that FREQ's platform is broadly applicable across a wide range of diseases with unmet medical needs where as OTIC has painted themselves into a corner by only focusing on the ear. Plus, FX-322 is barely touching the tip of the cochlea, maybe 15% to 20% of it at best and we are already seeing up to 40% response rate in particular etiologies. So even if the results aren't mind blowing, it is still a giant proof of concept for their technology so I can only imagine what they may be able to achieve once they can get deeper penetration into the cochlea with FX-345 and beyond.

Serious question, does OTIC have a patent on BDNF injection into the ear? Because I truly don't know if they do or not. I know FREQ has patents on not only the compounds they are using but the pathways they are activating as well. Does OTIC have patents preventing others from trying to put BDNF in the middle ear?

 

The responder definitions are not the same, Frequency Therapeutics uses the Thornton Raffin scale to classify a responder which is a pretty high bar. If you have a reliable history of word recognition scores and end up being a responder in that scale, that's a very significant response.

 

What are the chances of this happening? Do you think the FDA would be convinced to take that gamble?

 

@James Connelly, I can't see it happening. It would need to show not just statistically significant results but also high clinical benefit. Real world patient improvement that is noticed on a day to day basis.

 

What gamble does the FDA have in this? They don’t make any money. Their job is to determine if it’s safe and if it provides any improvements. Seeing how there is no other treatment for hearing loss, it will get approved if the clinical trial shows it has any clinical improvements and it’s safe.

 

Really depends on the closed door conversations the company has with the FDA. It was always considered the tradition that you needed 2 pivotal trials in order to file for NDA but this study looking back over the last 25 years found that approx only 53% of new drugs approved by FDA have 2 pivotal trials meaning almost half of all new approved drugs have only 1 pivotal trial.

"Between 1995 to 2017, the proportion of new drugs and biologics approved by the FDA using 2 pivotal trials declined from 81% to 53%. This trend in the decline of using 2 pivotal trials as the basis for approval is not unique to the FDA. A 2019 study documented the approval of 23 novel therapeutic drugs between 2012 and 2016 by both the FDA and the EMA based on a single pivotal trial."​

Based on the fact FREQ has run so many smaller trials with a combined dosed population getting close to 250 to 300 subjects now, they may be able to apply for NDA with only this 1 pivotal trial although I can't say with any percentage certainty how likely that may be to occur.

In regards to it being a "gamble" for the FDA, their biggest concern is safety first, then efficacy. If FX-322 continues to show a good safety profile, it will only need to barely show statistically significant improvements and nothing more for it to easily get approved. Even then, the FDA has been known to approve compounds with very little efficacy whatsoever when it occurs in a space with no other current treatment options as long as it is safe.

 

Sounds nuts to me quite honestly. Statistically significant says simply that there is an effect - possibly minuscule. It doesn't say that the effect is such that it's worthwhile pursuing. They're going to approve a treatment that involves punching holes through membranes for little demonstrated clinical improvement? You seem like you're across this stuff and I'm definitely not, so I'm not going to try to argue. Personally, I'd want to know that it actually works in a real-world way i.e. that it would improve my life.

 

@d'Wooluf, this is what I was alluding to. I see a minuscule signal to show it does 'something' positive. I can accept that. However, like you said, even if it gets approved, it's got to be sellable. In Britain the NHS approve and buy in treatments based on efficacy. Is it cost effective? They would not buy this on the results seen so far.

Lenire is approved as a medical device in Europe but the NHS has not bought it for clinics because its effect is non-existent to minuscule for most people.

How can Frequency Therapeutics sell it to health boards based on the results we have seen so far?

The average punter doesn't think like us. They have hearing loss on an audiogram and want to see an improvement or have a treatment that makes them hear better on a day to day basis, not a speech-in-noise test. They would need to experience a significant improvement before having needles stuck in their eardrums...

 

You're losing the forest for the trees and you're thinking like a tinnitus sufferer rather than a hearing loss sufferer. You have to look at the totality of all the data to see where the value in FX-322 lies.

To start, FREQ was told by the FDA that they would need to conduct a number of probing studies into all the different etiologies & severities of hearing loss if they wanted to apply for the broadest indication possible for FX-322 and that's what FREQ did. Now that FREQ has 5 studies under their belt, they know who FX-322 is most effective on and where they have seen the greatest response which is in the moderately-severe population where they have seen up to a 40% response rate and this is the population they are loading the Phase 2b trial up with right now. It was their first 2 studies in which they saw a consistent response in similar etiologies & severities but since then, they have ran an additional 3 more studies in other populations that have not shown a robust response so many have written them off entirely, completely forgetting about their early results. Another thing to remember is it's not going to work for everybody due to heterogeneity of hearing loss combined with the fact that FX-322 is barely touching the tip of the cochlea as you can see on slide 7 of their current investor slide deck.

You must also understand the test FREQ is using to determine statistically significant responders. They are using a 50 word recognition test modeled on Thorton & Raffin which defines statistically significant responders as improving by 10 words or more out of 50. It is generally agreed upon in the industry that a score of 20 words or less makes someone a candidate for a cochlear implant surgery which costs approx $30,000 to $50,000 before insurance. I've attached a graph showing the improvements subjects saw in their Phase 1/2 trial. Before receiving FX-322, both subject 1 and subject 3 qualified for a cochlear implant but after FX-322 administration, they no longer need a $30,000 surgery and can get by with hearing aids which only cost a couple thousand dollars. Subject 4 went from needing hearing aids to potentially being able to get by without them. I can guarantee an insurance company will be happy to shell out a few thousand dollars to pay for FX-322 for the chance that they don't have to shell out $50,000+ for a cochlear implant surgery.

The average hearing loss patient doesn't walk into their audiologist saying "I think I've lost 15 dB in my 4000 Hz range", most people go to their audiologist saying "I can't hear or understand the people around me". There are 2 parts of hearing; loudness and clarity. You say people only care about audiogram improvements so the best analogy I can think of to put it in perspective is, let's pretend you have a quiet, blown out speaker and you bring it in to get repaired and you tell them you want it louder but don't really care all that much about clarity. So the technician fixes your speaker and now you can crank it super loud but it still sounds blown out and garbled at a loud volume, are you going to be happy and satisfied with that?

I find it odd that you don't consider a speech-in-noise test improvement validation that someone can hear better on a day-to-day basis. There are people with hearing loss who use hearing aids that are still unable to go out to a loud crowded restaurant because they can't hear the people around them because they can't understand speech-in-noise. Even though some may think the results from FX-322 are disappointing, I can guarantee you it is life changing results for those that have been statistically significant responders.

To circle back to FX-322's lack of cochlear penetration, the fact we are seeing at least some responders is validation that FREQ is targeting the right pathway in the ear and bodes very well for FX-345 given its predicted cochlear penetration.

The last thing to consider is that current audiologic exams may not be capturing all the improvements patients are seeing. I recall a couple of years ago that someone shared that they were in one of the FX-322 trials and they said that they could tell the direction that sound was coming from better after participating in the trial and that is a perfect example of something that can't be captured with current exams right now. That is the reason the company has invented RADIAL, a patient report outcome which is a 40+ question survey asking patients how their hearing affects their daily life and is another way they may be able to capture improvements in order to get reimbursement from insurance companies.

If you don't see the value in a compound that can prevent people from going through an invasive, $30,000+ surgery, then there isn't much more I can say to change your mind.

 

Secondary outcomes of the current FX-322 Phase 2B should inform the decision on the question of "Do patients actually get helped." or "Does their quality of life improve in a meaningful way?"

Aside from the primary endpoint, which will need to be both CLINICALLY and STATISTICALLY significant based on current accepted testing measures for hearing specifically; FDA will also take into account the following to approve:

1. Research Assessment on Discernment, Intelligibility, Audibility and Quality of Life (RADIAL) Questionnaire [ Time Frame: Day 1 and Day 90 ]
   This questionnaire is a patient reported outcome measure.
   
   
2. Patient Global Impression of Change (PGI-C) Hearing Loss Scale [ Time Frame: Day 90 ]
   This questionnaire is a patient reported outcome measure.
   
   
3. Patient Global Impression of Change (PGI-C) Daily Impacts Scale [ Time Frame: Day 90 ]
   This questionnaire is a patient reported outcome measure.

 

Trial enrollment completion was announced this morning. With 90 days to go for the last subject + another 30 days or so for them to crunch the numbers, we should see a readout in early to mid February.

Frequency Therapeutics Completes Enrollment of Phase 2b Study of FX-322 for the Treatment of Sensorineural Hearing Loss

 

Interesting to see that they're testing effects on tinnitus too.

I like the part about how they tailored enrollment for this study based on how the previous studies turned out, as you mentioned before. Raises the chances of it being successful. Excited to see the results, yet prepared to be disappointed.

 

Looks like they over-recruited. Hoping that's a good sign for data outputs.

 

I really hope they calibrated their trial right this time! Praying for good results and a move to Phase 3, but I can't help but fear this'll end up like OTO-313...

 

They've done about all they can to make it as reliable a study as possible. To gain admittance you need an audiogram that is at least 6 months old plus 3 consistent testing sessions and on top of that they are audio/video recording every test session in order to have 2nd audiologists review for consistency.

Will hopefully help power the study more than initially planned. If they can achieve breakthrough designation, there could be a chance that FDA would encourage them to apply for NDA and use this FX-322-208 trial as their pivotal trial.

Yea, I think it gets overlooked that the only data they had to build their Phase 2a trial on was based on their small Phase 1/2 trial. In designing this Phase 2b trial, they have data from 5 different trials they used.

 

Anyone notice that the original amount of participants that were expected in this trial were 124 but now it’s 142.

What is everyone thinking? Do you reckon the extra participants added to this trial mean that this clinical trial has done well or not?

I read the slides about the potential for the FDA to accept the current trial as pivotal if the results are good.

In the slides it was stated that they would need to do an additional study after the pivotal phase. How long will that additional study be?

 

Do you still think there's a chance they will be able to skip Phase 3 and potentially go to market next year?

 

There still is a chance but for how big or little it is, I don't know. I don't think they'd get to skip it completely but they could end up applying for NDA and running a Phase 3 in parallel with selling it on the open market kind of like what happened with COVID-19 vaccines.

 

It's a guessing game as to what adding additional participants could mean. Maybe they have low confidence so they wanted to add more participants for a better chance of seeing some statistically significant improvements. Or maybe they had extremely high demand for the trial and they could fit in 18 additional participants with little effect on their current timelines so why not. Or maybe they are looking to beef up the trial in hopes of applying for NDA with the FDA with this being their one and only single pivotal trial. No one knows.

For Phase 3 I'd anticipate maybe up to 2 years, seeing that this Phase 2 is taking a year and 4 months for a readout.

 

Any chance if they make it to the final Phase 3 that they would lower or change their inclusion criteria? I know FX-322 focuses on the extended super-high frequencies (10,000 to 20,000 Hz) and I figure an extended audiogram would suffice, but they preferred volunteers with mild to severe hearing loss in the regularly tested 500 - 4000 Hz range where I assume the compound doesn't reach. Of course, I'm not going to DARE jeopardize the results of the trial if I'm not a good candidate.

I'm actually shocked about the performance of FX-322 with word recognition since most speech timbre falls below 10,000 Hz. This gives me huge hope for FX-345.