Frequency Therapeutics — Hearing Loss Regeneration

[Ed209]

That's *such* a misleading metric and I get triggered when people quote it, lol.

That's across all biotech and includes drug indications that have less than a handful of approvals in 10 years (e.g., Alzheimer's drugs). While other classes of drugs are much more successful.

There is literally no value in using an industry drug average to assign a probability on a brand new platform with no prior data.

I realise this and my original post actually mentioned how various drugs have very different statistical outcomes, but I deleted it because I sounded like I was rambling on. I know quite a bit about how drugs make their way to market, so I apologise for simplifying it to make my post brief. I just don't have the energy anymore after COVID-19. My brain is frazzled.

 

[Zugzug]

I haven't had the time to analyse the available data as I once would have done, but it's interesting to see a (sort of) optimistic outlook on here. I've only read a few of the posts, but it seems most are more positive than not.

My take on it is that it has to offer a beneficial therapeutic outcome. It could objectively work, but to what degree? How much will it change people's lives? A small increase in hearing that is either not noticeable, or that is negligible to the patient, would be disappointing. But, it could be the starting point of something bigger if there are ways to refine it. A complete failure to work on any level would be a disaster.

It will be interesting to see how it affects those tinnitus.

We've beaten this to death on this thread, but a big part of the efficacy results will rest on successful recruiting of people like the Phase 1 responders. As I showed, the people who respond tend to respond big time. Our expectations for below 8 kHz ranges should be low (although some did reach 6 kHz), but above that, it's quite promising, considering there is currently nothing.

@Diesel just made a good point, which is that making it through the finish line requires efficacy and success under rigorous safety requirements.

If it fails, I definitely don't think efficacy will be the reason. Everything is so solid. Even their pharmacology study demonstrates the drug is active in the cochlea in the above 8 kHz ranges. Then there's the ex vivo evidence of it working on human cochlea, as well as curing hearing loss in mice.

After analyzing the results compared to company messaging, I don't think the company is over-hyping efficacy at all. Probably under-hyping, actually.

 

[Diesel]

After analyzing the results compared to company messaging, I don't think the company is over-hyping efficacy at all. Probably under-hyping, actually.

Just to note... the hyping on Twitter or elsewhere doesn't affect whether it fails or not from an FDA standpoint. For the most part the FDA cares about significant primary outcomes + safety.

Any hyping that is misleading to the patient or doctor may result in business-level challenges that keep the drug from being distributed/sold. Or end up with an SEC investigation. So far, zero signs of that; they're playing it conservatively.

 

[Diesel]

We've beaten this to death on this thread, but a big part of the efficacy results will rest on successful recruiting of people like the Phase 1 responders. As I showed, the people who respond tend to respond big time. Our expectations for below 8 kHz ranges should be low (although some did reach 6 kHz), but above that, it's quite promising, considering there is currently nothing.

@Diesel just made a good point, which is that making it through the finish line requires efficacy and success under rigorous safety requirements.

If it fails, I definitely don't think efficacy will be the reason. Everything is so solid. Even their pharmacology study demonstrates the drug is active in the cochlea in the above 8 kHz ranges. Then there's the ex vivo evidence of it working on human cochlea, as well as curing hearing loss in mice.

After analyzing the results compared to company messaging, I don't think the company is over-hyping efficacy at all. Probably under-hyping, actually.

Should the Phase 2A outcomes (primary and experimental) look even remotely favorable, I anticipate that Frequency Therapeutics will expand the breadth of hearing loss topics / media / communications to include those new endpoints.

Let's not forget the Age-Related Phase 1B is due to end soon. So there may be additional focus on age-related hearing loss as well, should the outcomes be favorable.

I strongly expect them to focus on quality of life measures on Twitter big time in 2021.

 

[Guywithapug]

Everyone, I have been so inspired here lately reading all of the hard work put forth by such intelligent individuals here. The models, calculations and detailed data analysis provide such great insight into not only the drug, but also the potential efficacy. I know a lot is riding on a successful outcome so I decided to give it a stab myself. I have taken into account all of the data provided thus far to create my own efficacy projection below. Behold;

 

[Diesel]

Everyone, I have been so inspired here lately reading all of the hard work put forth by such intelligent individuals here. The models, calculations and detailed data analysis provide such great insight into not only the drug, but also the potential efficacy. I know a lot is riding on a successful outcome so I decided to give it a stab myself. I have taken into account all of the data provided thus far to create my own efficacy projection below. Behold;

View attachment 44138

I'm detecting a Christmas theme. We should analyze that.

 

[Keith Handy]

Everyone, I have been so inspired here lately reading all of the hard work put forth by such intelligent individuals here. The models, calculations and detailed data analysis provide such great insight into not only the drug, but also the potential efficacy. I know a lot is riding on a successful outcome so I decided to give it a stab myself. I have taken into account all of the data provided thus far to create my own efficacy projection below. Behold;

View attachment 44138

I think one of your p-values is off. It happens to the best of us.

 

[FGG]

Everyone, I have been so inspired here lately reading all of the hard work put forth by such intelligent individuals here. The models, calculations and detailed data analysis provide such great insight into not only the drug, but also the potential efficacy. I know a lot is riding on a successful outcome so I decided to give it a stab myself. I have taken into account all of the data provided thus far to create my own efficacy projection below. Behold;

View attachment 44138

Hmm... I ran my own Xmas themed analysis and I came up with these percentages for heading to the next phase at least:

 

[Zugzug]

Everyone, I have been so inspired here lately reading all of the hard work put forth by such intelligent individuals here. The models, calculations and detailed data analysis provide such great insight into not only the drug, but also the potential efficacy. I know a lot is riding on a successful outcome so I decided to give it a stab myself. I have taken into account all of the data provided thus far to create my own efficacy projection below. Behold;

View attachment 44138

If I stare at the pie with my left eye on the red half and my right eye on the green half, the right side of my brain is perceiving red, which is aggressive, and the left side of my brain is perceiving green, which is bright.

The left side of the brain is used for science and math. This medicine makes me feel positivity so the logical part of my brain is seeing the science the way I want to see it.

The right side of my brain is used for creativity and arts. I am thinking of creative and aggressive ways to act out my frustration if the drug fails.

@Guywithapug generates a sample size of n=1 for studying this pattern. Let's consider p, the percentage of people that choose to orient the pie with green on the right and red on the left. The null hypothesis is that p=1. The alternative is that p is less than 1. Under the null hypothesis, the p-value is the probability of achieving a result at least as extreme in the direction of the alternative hypothesis. Hence, p= P(X<=1)=P(X=0)+P(X=1)=1. With a p-value of 1, we fail to reject the null hypothesis that the true proportion of people who orient the pie in this way is 1.

We could be making a type II error, but this is okay, and is preferred to a type I error.

This result gives me reason to believe that a Phase 2a efficacy would demonstrate that most people view the pie similarly to @Guywithapug, which means that science is looking brighter.

 

[Jim Vasilakis]

So there is no chance that FX-322 will help people with hearing loss below the 8 kHz range? I really don't want to think that I'm going to spend the rest of my life with tinnitus!

I also wanted to ask what are the chances that tinnitus is caused by a 5 dB hearing loss at 4 kHz?

 

[Guywithapug]

I'm detecting a Christmas theme. We should analyze that.

Agreed! I think there is much left on the table to analyze!

Hmm... I ran my own Xmas themed analysis and I came up with these percentages for heading to the next phase at least:

Excellent effort and very inspiring to see we will proceed regardless of outcome!

If I stare at the pie with my left eye on the red half and my right eye on the green half, the right side of my brain is perceiving red, which is aggressive, and the left side of my brain is perceiving green, which is bright.

The left side of the brain is used for science and math. This medicine makes me feel positivity so the logical part of my brain is seeing the science the way I want to see it.

The right side of my brain is used for creativity and arts. I am thinking of creative and aggressive ways to act out my frustration if the drug fails.

@Guywithapug generates a sample size of n=1 for studying this pattern. Let's consider p, the percentage of people that choose to orient the pie with green on the right and red on the left. The null hypothesis is that p=1. The alternative is that p is less than 1. Under the null hypothesis, the p-value is the probability of achieving a result at least as extreme in the direction of the alternative hypothesis. Hence, p= P(X<=1)=P(X=0)+P(X=1)=1. With a p-value of 1, we fail to reject the null hypothesis that the true proportion of people who orient the pie in this way is 1.

We could be making a type II error, but this is okay, and is preferred to a type I error.

This result gives me reason to believe that a Phase 2a efficacy would demonstrate that most people view the pie similarly to @Guywithapug, which means that science is looking brighter.

Terrific analysis and review, it is heartwarming to see the support and that the science is behind us on this one!

 

[Diesel]

So there is no chance that FX-322 will help people with hearing loss below the 8 kHz range? I really don't want to think that I'm going to spend the rest of my life with tinnitus!

I also wanted to ask what are the chances that tinnitus is caused by a 5 dB hearing loss at 4 kHz?

Nobody knows for sure, yet. Frequency Therapeutics appears to be actively working on improving delivery — which many think means getting FX-322 deeper into the cochlea. So, it may not in the first product, but it's likely the next version will go deeper.

 

[Chetan]

So there is no chance that FX-322 will help people with hearing loss below the 8 kHz range? I really don't want to think that I'm gonna spend the rest of my life with T!!!

I also wanted to ask what are the chances that tinnitus is caused by a 5 dB hearing loss at 4 kHz?

I think if Phase 2a succeeds, eventually they can figure out how to penetrate deeper. All we need is confirmation the drug works in a meaningful way.

It's a drug delivery problem, not an efficacy problem that prevents lower frequencies. This can be solved. For all we know, Frequency Therapeutics buys Otonomy for their better drug delivery IP, or just reinvents it (given Carl LeBel was CSO at Otonomy for 7 years and likely knows what they need).

So don't lose hope!

For folks that are trying to estimate probability of success for Phase 2a:

I didn't go deep but I'm estimating it at 30-40% probability of failure, where failure is defined as there is little to no benefit in ANY of the cohorts that is statistically significant over placebo, for any of the key metrics of WR quiet/WIN/audiometry.

The source of this calculation from me is based on the average advancement rate for various different disease segments in Phase 2 when Phase 1 demonstrated efficacy and decent AE profile. I looked through, I doubt the AE profile will shift dramatically enough to prevent advancement or FDA denial, most issues were transient and minor/low grade events.

The real risk as we've hammered and @Zugzug also reiterated, are two things:

1. How likely is it they were able to recruit a more balanced group that matches the characteristics of their responders from Phase 1b?
2. How likely is it that those patients would demonstrate EHF hair cell LOSS and not just DAMAGE? (since we know FX-322 cannot repair damage, it needs the hair cell to be completely gone to regenerate the hair cell and result in synaptopathy damage being repaired with the regenerated hair cell).

These in my head, are the two primary risks.

For a little bit, biotwitter's concerns on placebo had me concerned a third risk was placebo response if you get placebo without ceiling effect issues, but looking at OTIC data and our detailed placebo data in patent has assuaged my concern (I'd already heard the KOL feedback but was surprised to see even transient placebo responders given test re-test variability is only 6% and not enough to trigger "clinically meaningful" 10% improvement in WR etc).

These are in my opinion the primary risks of Phase 2a not demonstrating any efficacy, or at least not statistically significant which is what they really need to advance to next trial stage.

I think as others have said, the odds of risk of failure to demonstrate this as a miracle cure are much higher failure risk just since that's still got numerous hurdles to clear.

When exactly is the Phase 1b age-related hearing loss readout anticipated? I'm actually most excited for that one given I expect age related hearing loss is likely the sub population along with severe that has the most hair cell loss and therefore stands to benefit the most (while also having least placebo effect risks).

I recall Q2 but was not sure early or late Q2?

 

[Koz]

This is really some thread now. We got hypothesis in abundance, repeated questions, doomsayers, optimistists, stock market "experts", reassurance seekers and memes for good measure.

Will I still be here for the next 2-3 years until this medication comes out? Probably.

Like a girl/guy you know you should really let go of but you just can't.

 

[spacehunter]

So there is no chance that FX-322 will help people with hearing loss below the 8 kHz range? I really don't want to think that I'm gonna spend the rest of my life with T!!!

I also wanted to ask what are the chances that tinnitus is caused by a 5 dB hearing loss at 4 kHz?

If your lowest dip in standard audiogram is only 5 dB then probably this won't be the cause. Have you checked your higher frequencies (above 8 kHz?). You might have there much more hearing loss without even knowing.

Regarding whether FX-322 will help below 8 kHz: We do not know so far and that's partially the reason why everyone is so excited to see the data from Phase 2a.

Even if FX-322 might not help you, there are other medication in the pipeline to fix synaptic hearing loss (not sure whether right expression), which is believed to be a possible reason for tinnitus. Possible treatments in the future might be OTO-413, maybe NHPN-1010 (Hough Ear Institute bomb blast pill) and PIPE-505 (?).

I know how you feel and can tell you we are all in this together. Hopefully research is doing more progress in this decade than in the last ones.

 

[paul mclean]

Only 10 more possible days for the 90 day data to be read out.

Cmon already, Frequency Therapeutics.

 

[scotty03874]

So there is no chance that FX-322 will help people with hearing loss below the 8 kHz range? I really don't want to think that I'm gonna spend the rest of my life with tinnitus!!!

I also wanted to ask what are the chances that tinnitus is caused by a 5 dB hearing loss at 4 kHz?

Very low chances of 5 dB causing your tinnitus...

 

[FGG]

So there is no chance that FX-322 will help people with hearing loss below the 8 kHz range? I really don't want to think that I'm gonna spend the rest of my life with T!!!

I also wanted to ask what are the chances that tinnitus is caused by a 5 dB hearing loss at 4 kHz?

We will have the penetrance data by the end of this month, likely by the end of next week.

 

[FGG]

I think if Phase 2a succeeds, eventually they can figure out how to penetrate deeper. All we need is confirmation the drug works in a meaningful way.

It's a drug delivery problem, not an efficacy problem that prevents lower frequencies. This can be solved. For all we know, Frequency Therapeutics buys Otonomy for their better drug delivery IP, or just reinvents it (given Carl LeBel was CSO at Otonomy for 7 years and likely knows what they need).

So don't lose hope!

For folks that are trying to estimate probability of success for Phase 2a:

I didn't go deep but I'm estimating it at 30-40% probability of failure, where failure is defined as there is little to no benefit in ANY of the cohorts that is statistically significant over placebo, for any of the key metrics of WR quiet/WIN/audiometry.

The source of this calculation from me is based on the average advancement rate for various different disease segments in Phase 2 when Phase 1 demonstrated efficacy and decent AE profile. I looked through, I doubt the AE profile will shift dramatically enough to prevent advancement or FDA denial, most issues were transient and minor/low grade events.

The real risk as we've hammered and @Zugzug also reiterated, are two things:

1. How likely is it they were able to recruit a more balanced group that matches the characteristics of their responders from Phase 1b?
2. How likely is it that those patients would demonstrate EHF hair cell LOSS and not just DAMAGE? (since we know FX-322 cannot repair damage, it needs the hair cell to be completely gone to regenerate the hair cell and result in synaptopathy damage being repaired with the regenerated hair cell).

These in my head, are the two primary risks.

For a little bit, biotwitter's concerns on placebo had me concerned a third risk was placebo response if you get placebo without ceiling effect issues, but looking at OTIC data and our detailed placebo data in patent has assuaged my concern (I'd already heard the KOL feedback but was surprised to see even transient placebo responders given test re-test variability is only 6% and not enough to trigger "clinically meaningful" 10% improvement in WR etc).

These are in my opinion the primary risks of Phase 2a not demonstrating any efficacy, or at least not statistically significant which is what they really need to advance to next trial stage.

I think as others have said, the odds of risk of failure to demonstrate this as a miracle cure are much higher failure risk just since that's still got numerous hurdles to clear.

When exactly is the Phase 1b age-related hearing loss readout anticipated? I'm actually most excited for that one given I expect age related hearing loss is likely the sub population along with severe that has the most hair cell loss and therefore stands to benefit the most (while also having least placebo effect risks).

I recall Q2 but was not sure early or late Q2?

I'm glad you brought up Carl LeBel and his Otonomy connection. Regardless of anything else, Otonomy knows intratympanic formulations probably better than anyone else...

 

[Diesel]

1. How likely is it they were able to recruit a more balanced group that matches the characteristics of their responders from Phase 1b?

Highly likely. Carl LeBel and Chris Loose have also said that the goal for recruiting in the Phase 2A was to create a "homogeneous" group of participants that "closely matched the conditions of those responders from the Phase 1/2"... LeBel in January also reiterated that outside of the Phase 2A inclusion criteria, Frequency Therapeutics had a "second filter" that they were keeping secret to again, ensure patients met those responder conditions. I stated a few posts back what I believed they were based on the available data from the patent/Phase 1b study and comments from the firm.

2. How likely is it that those patients would demonstrate EHF hair cell LOSS and not just DAMAGE? (since we know FX-322 cannot repair damage, it needs the hair cell to be completely gone to regenerate the hair cell and result in synaptopathy damage being repaired with the regenerated hair cell).

Frequency Therapeutics has referenced the terms, "Dead, Missing, and Damaged" as it relates to FX-322 regenerating inner and outer hair cells. The extent of the damage is unclear at the cellular level.

 

[Chetan]

Highly likely. Carl LeBel and Chris Loose have also said that the goal for recruiting in the Phase 2A was to create a "homogeneous" group of participants that "closely matched the conditions of those responders from the Phase 1/2"... LeBel in January also reiterated that outside of the Phase 2A inclusion criteria, Frequency Therapeutics had a "second filter" that they were keeping secret to again, ensure patients met those responder conditions. I stated a few posts back what I believed they were based on the available data from the patent/Phase 1b study and comments from the firm.

Frequency Therapeutics has referenced the terms, "Dead, Missing, and Damaged" as it relates to FX-322 regenerating inner and outer hair cells. The extent of the damage is unclear at the cellular level.

Yep I saw that, I agree FWIW with your WR deficit range, based on Raffin table for statistical significance ranges + clinically meaningful I expect WR deficit range minimum is likely between 70-85% minimum, WR deficit needed to be eligible. If they did do long term onset hearing loss criteria in the trial, I'll be excited since that does appear to reflect the responder population much more as well as durability too (potentially reflecting their damage is at a steady state vs patients who are earlier onset and may suffer ongoing lower frequency damage despite restoration of EHF ranges).

 

[Diesel]

This tweet sure could use some "educating".

 

[Zugzug]

Firstly, I'm really glad you are contributing. Investor chat is like a foreign language to me so I appreciate the posts to learn from.

2. How likely is it that those patients would demonstrate EHF hair cell LOSS and not just DAMAGE? (since we know FX-322 cannot repair damage, it needs the hair cell to be completely gone to regenerate the hair cell and result in synaptopathy damage being repaired with the regenerated hair cell).

FYI, the drug action (PCA) is activated for damaged hair cells as well. As @FGG pointed out to me, Carl LeBel said on the Tinnitus Talk Podcast that he didn't see the non-responding damaged cells as being a major issue. We'll see.

For a little bit, biotwitter's concerns on placebo had me concerned a third risk was placebo response if you get placebo without ceiling effect issues, but looking at OTIC data and our detailed placebo data in patent has assuaged my concern (I'd already heard the KOL feedback but was surprised to see even transient placebo responders given test re-test variability is only 6% and not enough to trigger "clinically meaningful" 10% improvement in WR etc).

I won't get into the weeds, but the placebo data was nothing at all. They set a 90% ceiling threshold before the study. Of the 4 eligible, 2 improved slightly and 2 worsened slightly by day 90.

Honestly, consistent with my philosophy for the belief that placebo effects are less with hearing, if anything I'm a lot more concerned about the fact that FX-322 treated Patient 932, from my p-value calculation post, had an improvement, but with a two-sided p-value of p=0.3273. This patient started at baseline at 8/50 (16%) so was far away from the ceiling effect. They improved at day 90 to 12/50 (24%), but it wasn't much.

Here's the thing though. Maybe they simply weren't a great responder. No drug is perfect. They still responded though and landed in the 6/15 responder group (>5 dB 8 kHz PTA, >10% WR and SIN). But I will concede that a Placebo patient with the same starting score could have had the same repeat WR improvement. But they also would have needed the PTA and SIN improvement by chance. Either way, I will concede that Patient 932 is a bad data point, and since I do science with my feelings, I will ignore it to believe what I want to believe.

Your point about duration is interesting, as Patient 932 had short duration (1.54 years, the shortest amongst responders).

 

[Tezcatlipoca]

This tweet sure could use some "educating".

View attachment 44145

He could use a talk with Toby.

Nothing screams "I know what I'm talking about" like polling people on Twitter.

 

[Gb3]

This tweet sure could use some "educating".

View attachment 44145

Toby 2.0.

 

[FGG]

This tweet sure could use some "educating".

View attachment 44145

The fact that they are calling this a "stem cell treatment" shows their level of knowledge about the drug...

 

[Keith Handy]

The fact that they are calling this a "stem cell treatment" shows their level of knowledge about the drug...

I expect "stem cell fraud" to be the next running joke in here.

 

[Diesel]

The fact that they are calling this a "stem cell treatment" shows their level of knowledge about the drug...

I recall an article a while back confusing FX-322 as a "stem cell" treatment. I think the author mis-understood Frequency Therapeutics' description of progenitors being "stem cell like." I imagine it's happened more than once, since it's pretty difficult to get someone's head wrapped around progenitors in less than 2 minutes. So, now every once in a while FX-322 is now erroneously associated by the uninitiated as a stem cell treatment.

 

[Aaron91]

He could use a talk with Toby.

Toby 2.0.

If you guys keep this up, Toby is gonna become a meme lol.

 

[FGG]

I recall an article a while back confusing FX-322 as a "stem cell" treatment. I think the author mis-understood Frequency Therapeutics' description of progenitors being "stem cell like." I imagine it's happened more than once, since it's pretty difficult to get someone's head wrapped around progenitors in less than 2 minutes. So, now every once in a while FX-322 is now erroneously associated by the uninitiated as a stem cell treatment.

It's funny how that works. I spent a High School year in England and took A level history and one of the only things I remember is that millions of "witches" didn't die in England witch hunts (60,000 tops) but one history book got it wrong and everyone just quoted that one (and quoted others down the line) for years until someone realized the mistake.