Frequency Therapeutics — Hearing Loss Regeneration

Finally found out where to buy FREQ stock here in Slovakia, so I am going to buy.

 

I won't buy all now for a number of reasons including - need more data this time around, and stock market may see another pullback. Option players are smart. I wouldn't place all eggs in one basket.

There are many way better and safer investments.

 

Biotechs are risky, it's like a blackjack table, they post good results, you get handsomely rewarded. They post bad results, it stays a penny stock. My eggs are all spread around. It would hurt me if I lost $25,000. I have a very diverse portfolio.

 

Naaah bruh, I think Frequency Therapeutics will pass Phase 2b, those results are clear, problem is that it only hits up to 8 kHz, that is why it is not such a tremendous result, but in this Phase 2b, they test also up to 16 kHz (they didn't test previously) and there are many other tests that weren't done before + portfolio of subjects is much more precisely selected.

I am 100% sure it will work, and not just because I want it to work, but because the data is there.

 

The data isn't there though, which is why the stock has crashed 85% in the past year (and that's on top of the 80% it crashed when Phase 2a failed). They did 3 Phase 1b trials last year, and all of them had poor results.

• Phase 1b-111 Open Label Trial: 34% response rate for 10% or more improvement on WR scores over an untreated ear. No placebo group. Months later they retested many of the participants and found 4 new responders. However, in this release they say the total is now 9 responders out of 32, which would yield a 28% responder rate, and an original responder rate of 16% (someone please correct me if I'm messing something up here). Additionally 4 of the 5 original responders were no longer statistically significant (and the last one didn't come back for testing, so we don't know if they stayed statistically significant). So no single time point had more than 16% of patients counting as a responder. I don't know why they originally said there was a 34% response rate when they later claimed there were only 5/32 responders.
• Phase 1b-112 Age Related Hearing Loss Trial: "the Company observed no significant treatment effect with FX-322 administration compared to placebo".
• Phase 1b-113 Severe Hearing Loss Trial: The response rate was 16% for the FX-322 group and for the placebo group - they did not beat placebo. Additionally, the response was for a sound-in-noise test, there wasn't a response for a words-in-quiet test (which is what they had been previously touting).

Even more damning is that they aren't seeing high frequency audiometric improvements, even though they tested for them in Phase 2a.

Compare all of this to Otonomy's OTO-413. In their latest study the response rate was ~40% (and that's for 2 time points instead of 1) and the placebo response rate was ~13%. That tells me that a placebo rate of 15% isn't unreasonable. Given FX-322's track record, I don't see how they can beat a placebo rate of 15%. I certainly hope that it passes and they can continue on with their science, but don't be fooled - they have an uphill battle in-front of them.

 

@Geatly, the results are clear? The results are murky at best.

 

There was something funny with one of their trials if I recall correctly. It warped the results I believe.

 

I think the point being made is that it's getting harder for them to argue that the reason for FX-322 underperforming is trial design (rather than the drug itself) when the results have been lackluster in multiple trials.

I just wish we had better ways of diagnosing the inner ear. Is the problem that FX-322 doesn't propagate enough? Or does it propagate well enough but it just doesn't work?

I do hope that they iron out these kinks, wherever the problem may lie.

 

I thought they didn't test above 8 kHz in previous clinical trials, but it seems like they did, which is quite depressing for a person like me. If it is going to fail, that means nothing helpful is coming within the near few years, but thank you anyway for clarifying facts.

 

They did test with the extended high frequency in the Phase 1Bs and 2A. However, didn't see anything meaningful. The Phase 2A data was trashed by patient bias, and a loose entry threshold, so there isn't much to say about the Phase 2A. As for all of the varying Phase 1B outcomes, the underlying conditions for each Phase 1B were different, and the reason those exist are because the FDA compelled Frequency Therapeutics to test these different population groups. This is why they raised more funds in July/August of 2019. They've mentioned this several times, but there's a lot of cherry picking that happens here to make a "case" for/against the firm. Anyway, the Phase 1Bs didn't produce favorable outcomes for those specific underlying conditions, so the Phase 2B is focused on where they did see improvements on the original Phase 1/2. This is covered in their investor deck, and in a couple different videos since November of 2021.

In a recent interview, the SVP of Product, Kevin Franck actually acknowledged that they aren't seeing audiogram improvements, but have seen them in WR and WIN tests. My interpretation is that he implied that FX-322 is having an effect on regenerating Inner Hair Cells and not Out Hair Cells, or not enough to show improvement on the audiogram.

I would advise not stacking up the past "failures" to predict the Phase 2B will be failure, or that they have some "uphill battle." The uphill battle, IMO was the Phase 2A and the Phase 1Bs. Each trial was designed differently, so again, it's cherry picking to think this way.

 

I don't understand why it would regenerate inner hair cells and not outer. The whole purpose is to "activate" progenitor (supporting) cells in the cochlea, whether inner or outer. It seems these guys have no idea what they are doing. I hope I am wrong though.

 

For myself the more uncomfortable part of the current situation is waiting for the results for another month. The anticipation is hard to balance and the possibility of either a positive outcome or a negative outcome are heavy concerns to bear.

 

I am personally more interested in Inner Hair Cells than Outer Hair Cells as I have 5 dB across all frequencies tested up to 8 kHz, and my real problem is decreased clarity, which is far more important for me.

 

I have a hard time understanding this too. Wouldn't they have seen only IHCs regenerated when they did their pre-clinical work? I think they're trying to reconcile what they believe is a real signal (the improvement of WR scores in 16-33% of patients) with no measurable audiometric improvements. I've heard some speculate that maybe the shot "cleans out" the inner ear, which would explain the corresponding placebo response rate. It could also be that the new hair cells that are created are inferior to real ones.

A while back someone posted a study that was done on a drug that's similar to Audion's drug (which changes support cells into hair cells), and it showed that the new hair cells were "hair cell-like", but didn't function identically to real hair cells. Maybe FX-322 creates flawed hair cells that don't function exactly right but still help out in some aspects? This is just pure speculation though.

It's not cherry picking though. Frequency Therapeutics has done 5 studies: 201 (Initial Phase 1/2 study), 111, 112, 113, and 202 (Phase 2a). Phase 2a was botched and its data isn't any good. My post centered around their 3 most recent studies. The only study I left out was their original 2018 Phase 1/2 study which had 23 participants and showed a (then) promising 33% response rate for 10% or more in word recognition improvement. My post centered on their most recent data collection which represents the majority of their data.

They had 3 big failures last year, and the minor success of 111 is marred by the fact that there was no placebo group and no one who had a benefit at 90 days was shown to have maintained that benefit. Even Audion's drug showed efficacy signals that persisted after a year. Their uphill battle is proving that the signal they're seeing is real and producing a study that definitively shows it.

 

201-1 = (Phase 1/2) SSNHL and NIHL Study Mild-Moderately Severe [18-65 y/o]
201-2 = Phase 1/2 Durability (Long term study)
111 = Study of Administration Conditions - All Etiologies, Mild-Severe [18-65 y/o]
112 = Age-Related Hearing Loss Study - Mild-Moderately Severe [66-85 y/o]
113 = Severe Hearing Loss Study - All Etiologies [18-65 y/o]
208 = SSNHL and NIHL Study - Moderate - Moderately Severe [18-65 y/o] AKA Phase 2B

201-1 and 201-2 had arguably favorable outcomes. Same patient groups.

208 is the only one that builds on the 201-1/202-2 study.

The others were arguably different studies with different underlying conditions. One cannot merge them together and claim "it's uphill" when they don't build off of each other, except for 208. Which builds off the 201-1 / 201-2 trials that had positive outcomes.

 

Honest question, though: how is a lack of benefit against placebo across most of these trials anything less than a worrying sign? I understand the etiologies/patient populations are different, and I don't think it means that FX-322 is doomed or that they can't make it work eventually, but I don't see how it is anything other than a little concerning.

(It should go without saying but I'll say it outright, just in case, that I want this to work.)

 

A couple things to consider:

These studies are including the widest range of etiologies & severities possible in order for them to probe to find out who it does and doesn't work on so you have to do subset analysis to tease out the valuable data. The FDA told them they would have to do this in order to apply for the broadest indication possible for the drug.

For example, the Phase 1/2 study included at least 3 patients with completely normal hearing because the primary objective of these early trials is to test for safety and if a patient experienced a worsening of hearing after receiving the drug, that would be a huge safety concern so efficacy is just a secondary outcome measure they are looking at behind safety and why they are including patients that dilute the final numbers.

As for the 111 study, the company uses 2 different levels to define a "responder". The first level a responder can achieve is called "clinically significant" which is defined as a patient seeing a 10% absolute improvement on a 50 word test so an improvement of 5 words makes a patient a clinical responder and this amount of 10% is often used by clinicians as it is considered the level at which one may notice differences in their daily lives.

Next you have statistically significant responder which is an additional level of statistical rigor they apply that roughly doubles the bar subjects have to reach in order to be considered statistically significant. It's at this level that they can say with 95% confidence that improvements that patients are seeing is due completely to the drug and not placebo or chance.

So you are correct that 34% of subjects saw 10% or greater improvement but a subset of those subjects then went on to see statistically significant improvements, some nearly doubling their word scores. This is the equivalent of a subject going from being a cochlear implant candidate to possibly getting by with lifestyle modifications and over the counter hearing devices which has never been done before with a therapeutic in the hearing space. And it is also true that many patients see a waning effect over time but do not lose all the benefit. The company stated that those who fell out of statistical significance still had word scores that were well above their baseline scores but they no longer reached the bar for statistical significance. You can see this when you look at the bar chart from the Phase 1/2 study. The fact that this waning effect over time is consistent across multiple trials suggests it is a true hearing signal we are seeing and its not just a placebo (in my opinion).

Another thing to consider is that it is likely that these current exams aren't showing us all the different ways patients could potentially be improving. I remember it was shared in this thread a couple years ago that one trial participant reported being able to sense sound directionally better after the trial, if this is true and not just anecdotal hearsay, that would be a perfect example of an improvement a patient saw that was not captured by a WR test or audiogram and is why the company invented the patient reported outcome measure RADIAL which is over 40 questions asking patients how their hearing affects their daily life before and after the receive treatment.

If you look at the mouse cochlea image the company has shared, there was relatively few new outer hair cells at the highest frequencies after FX-322 treatment but at the lower frequencies it looks like the outer hair cells had a more robust response.

 

I don't see it as something that is "worrying"... I see it as scientific discovery happening in real-time, and in front of the public. This is probably why we see so many dissenting opinions here, because it isn't something the general public is all that familiar with. The FDA advised Frequency Therapeutics to test the drug for these other types of conditions and see if the drug provides a benefit. Turns out for 112 and 113, due to the severity of the damage in the cochlea (severe SNHL) or the underlying condition (age-related hearing loss), that the drug didn't have a statistically significant effect when compared to placebo. If I were a manager at Frequency Therapeutics, that's a sign to me to stop investing in that front; the FDA's recommendation didn't produce enough value / de-risk going in those directions to warrant further investigation.

Now, looking at the 201-1 / 201-2 studies, there were benefits observed against placebo (201-1) and when compared to the untreated ear (202-2). Furthermore, it appears based on those that got drug, they were able to identify a subset of patients that are more likely to produce an outcome that made them a "responder" based on certain improvement metrics. So, from that, they now have an "ideal patient" that is backed by statistics to respond favorably to the drug.

Let's take a pause and review what the FDA REQUIRES to approve a drug:

1. That it is safe, or safe enough that the benefits far outweigh the risks.
2. That the drug produces a significant outcome based on at least 1 accepted tests when compared to placebo.
3. A specific patient group can be reliably identified.

Now let's look at FX-322 when compared to those FDA approval requirements.

1. FX-322 has been well tolerated across all studies, with no serious adverse events. All studies. Safe. [Check]
2. FX-322 in the 201-1 / 201-2 studies produced a significant outcome on the Word Recognition Test. [Check]
3. The patient group that appears to be most reliable has been acquired and/or sudden SNHL at the Moderate-Moderately Severe hearing loss levels. [Check]

So, put that manager hat back on at Frequency Therapeutics. If they wanted the FDA to approve FX-322, so that it can get into the market, and raise cash that can be invested into making a better FX-3xx drug; it would be super smart to build a trial that hits the three points I outlined above.

That's the 208 study.

----

Personal note:

I've been fortunate enough in my career to see "behind the curtain" in product development / engineering for many products that everyone here reading this post has seen every day. A big part of all of those products, especially the ones that have been able to mature over many iterations, is the ability to discover and fail a few times. Sometimes, we're talking about hundreds of failures, Each failure has its lessons. This company/drug is following a process that is no different.

I read a lot of armchair quarterbacking here, and cherry picking of events / outcomes; of which mirror the same type of quarterbacking and cherry picking that I have seen for many of the well-known products from "household names" in other industries. And as someone who's sat in the meetings "behind the scenes" when decisions are being made as learnings are made, it's clear the outside world has no idea what's really going on.

None of us are there at Frequency Therapeutics as managers making the decisions in real-time based on these outcomes. I heard a very panicked management team on a call after the Phase 2A collapsed, and I have heard that before in product failures where the stakes are high. Many of us here got a glimpse "behind the curtain" and I think many opinions were formed. Personally, I think lessons were learned, and their decision to form the 208 study to build off of what they had understood to be the best outcome based on the FDA guidelines makes a lot of sense.

 

I can certainly agree with this, and I see this happening with a lot of other treatments here (most notably, recently, with Dr. Thanos Tzounopoulos' drug).

I fully expect the process of trialing and refining the drug to be long and tortuous. I think that's a sentiment we both share, we just seem to have different reactions to the tortuousness of it. Still really appreciate and value your input, as I know you've been following this and providing your insights for a long time now.

 

I'm honestly just glad they're still trying FX-322 out.

 

I've seen this image before but never looked that closely at the differences between regrowth re: inner and outer hair cells and at different frequencies.

They treated the mouse cochlea in a more direct manner than your typical IT injection, right? Would that change the outcome for outer hair cell regrowth in regards to position within the cochlea?

I find this a little confusing given that it's assumed that FX-322 works best at restoring the hair cells in the highest frequencies as it enters the cochlea. And indeed, that image represents it working very well on the inner hair cells from the very highest frequencies, but like you said, the outer hair cells don't really start showing up in numbers until a good way down the line.

Does anyone have any sort of scientific explanation for this? I'm pretty sure my EHF are shot, so I would have thought FX-322 would have been, theoretically at least, ideal for me. But now I'm not so sure.

 

Since the science is still unfolding in front of us, it's really a guessing game right now. I can't say I've stayed very up to date on the arguments of FX-322 preferring inner hair cells over outer hair cells but it looks like that may be the case with the little evidence we have seen so far, at least at the higher frequencies.

Maybe the shape of the cochlear spiral and the order in which the hair cells were grown in the womb plays an effect on which hair cells in which sections of the cochlea may respond to FX-322 better than others. Just my hypothesis.

And yes, I do believe they are able to dose the mouse cochlea with better distribution into the lower frequencies than they can in humans, the greatest limiting factor right now is the round window membrane and how high of potency is needed to pass through it and hence why FX-345 should help improve that.

 

I once talked to a researcher in hair cell regeneration and she said that, for the moment, the main problem when trying to regenerate human hair cells was the appearance of cancer cells as a result.

ENTs have several times told me that new hair cells would need to be connected in the proper way, each hair cell correctly connected, to the rest of the hearing system.

To me it seems incredibly complex and to be honest I do not think hair cell regeneration will be attained in our lifetime.

 

I disagree with you as I believe we are already seeing it happen.

For the longest time, stem cell transplants and gene therapies were thought to be the most likely solution for hearing loss and there is a greater risk of cancer with these approaches if say you don't quite transplant the stem cells to the proper location or your gene therapy activates targets it shouldn't. There is still some risk with this in FREQ's small molecule approach but it is a much smaller one since it is just a temporary activation of the cells that are already in place in the body and it is done using small molecules that are quickly cleared from the body within hours after administration.

In regards to complexity, FREQ's approach is actually quite simple. All they are doing is providing a temporary signal to the progenitor cell using small molecules to tell it to turn on and do the job it has already done in the past; replicate itself via mitosis. That is it. The supporting cell then goes through mitosis and we are left with a supporting cell and a fresh brand new hair cell that knows how to connect in to the auditory system on its own, forming hair bundles & functioning ion channels without needing further input. This is why I believe in FREQ's approach and why I think their method is superior to all their competitors.

 

Disagree. There are at least 6 companies working on different hair cell programs right now: Frequency Therapeutics, Otonomy, Decibel Therapeutics, Akouos, Hough Ear Institute and a not disclosed partner working with Astellas.

There are also many researchers in universities working on hearing loss restoration, with a focus on hair cell regeneration.

 

This is something that gives me hope, to be honest. Even better is that all these companies are, as far as I know, working on different approaches to hair cell regeneration, so even if one of them fails, there's still hope for the others.

 

Honestly there comes a point it turns into throwing enough stuff at the wall that eventually something will stick.

 

And where are the results?

 

This comment describes what should be known as the “Tinnitus Talk Paradox”...

1. A claim where ENTs have academic training, years of experience, and don’t know shit about things like tinnitus. Yet, are a reliable source of research in the space.

2. Anecdotes from a “researcher” broadly defines the outlook across the entire industry, without any data cited.

 

OTO-413 reported positive results not too long ago.