Frequency Therapeutics — Hearing Loss Regeneration

[RichieTheKid]

I'm still 100% supporting Frequency Therapeutics and willing to try it. Any chance of getting it in the UK later this year?

 

[Ed209]

I still feel dead after yesterday. What do we do now?

Nothing has changed.

People mustn't get too emotionally invested in this. See it for what it is until there's some clinical evidence to show that it works. Until then, treat it the same way as any other drug that's in development, and be brutally objective.

I really feel for all of you that are depressed by the results, but that's biotech for you. Expect the unexpected in either direction.

I wasn't impressed with the phase I/II results, but it was primarily a safety trial, so I didn't really read too much into it in terms of efficacy. I noticed the issue was raised about people faking their results, but there was talk of the WR scores being falsified back in the Phase I/II trial. I remember someone posting a video on how to "fake it." The 10 dB improvements that were seen in 4 people had no clinical value. I believe there was an increase seen at 8 kHz; my audiogram had the same increase at the same frequency and I did nothing other than take a few vitamins. This is when I first started to doubt how effective FX-322 might be, but it was only a small cohort. Before those results were published, everyone was hyper-bullish saying more doses will equate to better results, etc, etc. I thought people at the time were getting too far ahead of themselves. I also didn't like how they spun the success of those earlier results to the media before they showed any data. The headlines and sound bytes had everybody salivating.

The latest results seem to be more of the same. They should have doubled down on the trial design and made sure the WR scores couldn't be faked. What I ultimately wanted to see was the slightest hint of an improvement on the audiograms. I realise audiometry is a tad antiquated, and that there could well be an improvement in hidden hearing loss, but I think people generally expected far more than that; especially in the earlier days.

Sorry for sounding negative, because it may come across that way, but those are my honest thoughts and feelings.

 

[Zugzug]

I actually have a theory about this and that is that people with truly bad word scores likely had audiograms that would put them more in line with the severe group (makes sense as they hear worse) and they didn't get any real borderline cases this time. This would mean all the low word scores were potentially fake but ironically the fake ones were preferentially selected to match the criteria.

The actual severe group shouldn't have this problem.

I see your argument. I guess I'm really surprised that there aren't simple tests to see if the WR scores compared to the overall audiograms match.

As we've talked about a lot, this is challenging because there's OHC vs IHC loss. Word scores are a huge part of their marketing platform so they may not have had a huge problem with seeing low word scores that didn't quite match the audiogram because it is possible (several people on the forum have this issue, including yourself).

It looks like maybe they need to just keep it simple and look for people with evenly dispersed IHC and OHC loss. They could then reject people based on their scores not matching. Someone could still lie, but it's at least better.

Okay so overall, I take it as you think there were a lot of cheaters. In my analysis, the number of cheaters was assumed to be 10%. For the sake of argument, let's say it's very high -- like 50%.

If this was the case, even if the cheaters were proportional between groups, the "cancelling each other out" effect would be greater the closer that number gets to 100%.

So your design thesis is that they really screwed up the word screening. They had many applicants, tried so hard to get low word scores in order to match their marketing that they picked up a bunch of liars?

I guess I have doubts that the number of liars was that high. The individual data would really help answer this question. As I think we agree upon, people don't massively improve word scores (honestly) by chance. If the individual data shows a bunch of screening to baseline comparisons that look like the outlandish responders we saw in Phase 1, then the company may have a good argument.

I have doubts that this will be the case. I think it takes some serious guts to intentionally screw up a word, but maybe that's just me.

 

[FGG]

The craziest part of all of this is that someone definitely lied because people don't double word scores from placebo.

So it depends on which of the 3 you think did:

1) Frequency Therapeutics themselves
2) Phase 1 participants
3) Phase 2 participants.

Frequency Therapeutics says they have direct evidence it is option 3, which would mean the drug does work. Time will uncover this.

 

[Gb3]

I actually have a theory about this and that is that people with truly bad word scores likely had audiograms that would put them more in line with the severe group (makes sense as they hear worse) and they didn't get any real borderline cases this time. This would mean most all or all of the low word scores were potentially fake but ironically the fake ones were preferentially selected to match the criteria.

The actual severe group shouldn't have this problem.

So do we think people faked their audiograms?

 

[tomytl]

Everyone,

I'm a long time reader of this thread. I appreciate all of the wisdom found here - you guys are truly a great community.

Yesterday was awful for me. I couldn't get out of bed. My gut reaction to the perceived failure of FX-322 hit me hard. I've had so much hope hanging on this. I was also devastated at losing almost $40k and most of my stock account. (I am an idiot, wasn't diversified, not an experienced investor, and was emotionally invested, foolish I know.)

But today gives me some new hope. The hopeful and rational messages on the last several pages are encouraging and I believe we do have legitimate reasons to still have cautious faith. We aren't done yet with this story.

For those of you who are financially invested too and really down about it, just search Google for info on Oppenheimer's new price target of $20 and their reiterated "buy" rating as of late yesterday. Or B. Riley's revision to $35. Also, there are some good YouTube videos on price analysis and predictions too. It lowers my "freak out" levels just a bit. Might for you too.

I'm not going to post links here because I know it's not a financial board and I don't want to distract too much from the most important topic; efficacy. But just trying to spread some reasons for hope from the financial side if anyone needed that right now.

Anyway, even though it was a double gut punch yesterday, I think if we can stay calm and continue to be patient (it's so hard I know), we still have a chance for this to help us have better lives.

Don't blame yourself too much for your investment, it really looked so promising for many of us.

I just didn't invest, because I was too lazy to fill out all agreements to access NASDAQ from my platform.

But yes, it's a double strike in the wrong direction.

I really hope for you and other guys on this board that things might turn to the good again.

 

[Street Novelist]

I still feel dead after yesterday. What do we do now?

This drug has to work. We've invested far too much time for it not to. Hopefully it's because they didn't wait long enough between injections.

 

[Lucifer]

I wrote them an email asking what their next step would be.

What is Frequency Therapeutics' email?

 

[r34d1ng]

Or B. Riley's revision to $35.

Don't want to be mean, but B. Riley's $35 revision was prior to the Phase 2a results.

Correct me if I'm wrong, though.

 

[weab00]

This drug has to work. We've invested far too much time for it not to. Hopefully it's because they didn't wait long enough between injections.

Nothing has to happen, the universe doesn't owe us anything. Our suffering and emotional investment means nothing for any of these drugs.

 

[Zugzug]

The tl;dr of the situation is that ignoring the reasons why (lying, poor and inconsistent test administration pre-screening, etc.), the actual math going on is that all groups were shifted up closer to the ceiling effect in an unexpected way (compared to screening) at baseline.

Though for different reasons, this is not dissimilar to why OTIVIDEX failed. What likely happened (speculating) is that they legitimately had frequent vertigo attacks at screening so got in (compare to FX-322 Phase 2a participants having low word scores across all groups). Then entering the trial for OTIVIDEX alone probably initialized a small placebo effect. This is equivalent to many participants in FX-322 Phase 2a having higher baseline scores.

From there, placebo and treated groups improved, but they were working against a ceiling effect. The OTIVIDEX failure of design was almost surely innocent and a true placebo effect, whereas the FX-322 design could have been some dishonesty combined with standardization incompetency.

Either way, whether it be for moral reasons or design reasons, it does appear that any test involving the ears is not straightforward. Goes to show why most trials have such rigorous standards because even the unforeseen can happen.

 

[HootOwl]

We need matching "Team Goop" t-shirts. Our mascot is an injured but determined-looking bull.

Don't tempt me, I will waste money on this.

 

[Lucifer]

This drug has to work. We've invested far too much time for it not to. Hopefully it's because they didn't wait long enough between injections.

I really hope that's the case. I'm praying that the results are positive for age-related and severe hearing loss trials. I'm more worried about the severe category even though we have a positive anecdote saying that he had improvements in his word scores and 3 frequency bands and can also hear birds and sing and has no tinnitus 5 months after his last dose of FX-322.

I still believe FX-322 can go to the pivotal phase without severe hearing loss as long as their age-related hearing loss show improvements as well.

It depends if the FDA allow FX-322 to go to the pivotal phase without the severe category. They can always fix the delivery method after it's out in the market. FX-322 will still help millions of people in the mild-moderating hearing loss.

 

[Cernuto]

The placebos that showed improvement may not have been fakers. It would not surprise me at all if the qualified audiologists that conducted the tests did them poorly or perhaps even fraudulently.

 

[Aaron91]

Now, regarding the lying itself. Here's what I can't understand. So the theory is that lying got them in, but then their baselines were their real score (higher). In a sense, how does it matter how they got in if they still performed the study correctly?

@Zugzug, I believe this is not necessarily the case, but someone please correct me if I'm wrong. My understanding is that there was an initial screening test for all patients to determine whether you could participate in the trial or not. Once you had been screened (and selected), you then undertook the same tests again. These second tests were what were used as baseline - not the screening tests. What I find mind boggling is that in one scenario, these patients would have had to have lied twice: once during their screening and again during baseline. What I'm thinking though is, how likely is that given that once you know you've been admitted to the trial it doesn't matter how you perform on the new baseline test? What it really comes down to then is whether Frequency Therapeutics were aware of either scenario. The thing is, they weren't - they were blinded. It's not as if Frequency Therapeutics would have caught on to higher baselines that were different from the screening tests. So I can see one of the two possibilities here. Either patients lied/faked their results twice OR they faked their initial screening and then performed as normal, but by the time they performed their baseline Frequency Therapeutics were already blinded so it was out of their control. In the latter scenario, it's possible that patients who were originally thought to have word minimum WR deficits actually were way too close to the ceiling. In either scenario, Frequency Therapeutics would have only been able to figure out what was going on by cross-referencing against historical records. To put it simply, either scenario is a shitshow if I've ever seen one.

If I understand what LeBel is really suggesting, it's that the treatment group was actually held back because the baselines started closer to the ceiling effect than they anticipated, disproportionately to the placebos.

This would definitely appear to be part of the reason as to what went wrong. Coming back to your example, what if we imagined the complete opposite: that all or most patients lied/faked their results? In this case, the distribution would almost seem to not matter at all. Every patient in every group is unreliable, regardless of what the end result is. The whole trial would be worthless and the only way around it is to start again.

Btw, I'm loving the new avatar, @Zugzug. I may change mine to a bull eventually, but need to cool off first and figure out if I really think Frequency Therapeutics can turn this around.

 

[Drachen]

Just spent the last hour or so catching up on posts since the news dropped. I know I'm a bit late to the party, but I thought it was going to be this Thursday instead. I was utterly shocked when I checked $FREQ and saw it brought down to single-digits. The feeling in my chest echoed the sharp decline in the graph.

I am absolutely devastated by the news. I will admit that I perhaps had too high of expectations given what has been shared here in the thread, but even so I never would have predicted the (nearly) worst case scenario to unfold. Everything was on track. Everything looked right. It was simply only a matter of time, right?

Ever since I first being experiencing my symptoms, the one thing I found helped me the most was the prospect of regenerative medicine: real science that was producing real results. As bad as one day would get for me, I would just know deep down that I have to brave through it and I would no longer have to deal with this whenever I could just get these injections in the future. That was to be the end.

So far, 2021 has proven to be a complete bitch of a year. The news with respect to OTO-413 was already a blow, but this is something else. I do not believe that FX-322 is dead in the water, but my perspective has been altered dramatically, and I am nowhere near as optimistic as I was before. The best case scenario for the drug implies release will be even further out than it already was, and I know for many of you that is unacceptable.

Not really sure what I want to say at the end other than I am continuing to hope for the best for all of us, though that is a bit hypocritical to say considering at this point, the word "hope" leaves a nasty taste in my mouth. I want to believe that there will be a day in the relatively near future in which we all can stop suffering daily from such insidious conditions.

Sorry for the blogpost, but I needed a way to get my emotions out after all this. Best wishes, all.

 

[Tezcatlipoca]

Can't FREQ just continue? I don't understand the pessimism. They're constantly talking with the FDA about development. And they already have a couple of trials with positive results.

 

[FGG]

So do we think people faked their audiograms?

I'm not as sure about that but per Frequency Therapeutics, they say there is evidence they lied about word scores which would be a good explanation of the difference from Phase 1 so it's possible EHF numbers were purposely depressed on entry for at least some of the same group.

I don't think people could lie about their standard audiogram. They measured stability as having one from 6 months prior that matched the baseline they took that day. We can almost completely rule that out.

Almost no one has long term EHF results, though, and those could be faked to seem worse, with the goal of lowering your overall pure tone average (EHF was included here if you had it). I think if someone would lie about word score, they'd lie about any of it.

The other option is that since EHF audiograms are supposedly "hard to calibrate" for audiologists, it's possible there was variable between testing sites.

In my own case, i have 50-55 dB loss at 12 kHz (worse further up). That was my first measurement. I got it tested again at a different place 6 months or so later and got 40 dB. I had to beg that place into giving me that measurement. They didn't even want to set it up because "it's very hard to calibrate".

Anyway, I didn't notice a hearing improvement but I initially took it as a good sign until I got it tested again at the original place and got 55 dB.

I would *think* this would be less of a problem with a testing site than the regular ENT clinics but honestly I don't know anymore. I have so many different possible explanations for this that only time will uncover. Once they unblind patients they could assess that as well (assuming my issue wasn't just due to a particular incompetent audiologist and this is even a possibility in the testing conditions).

Here is another factor. Testing centers are impartial. They have to almost be like robots to remove "bias" and just follow instructions. So if you had a certain PTA going in and qualified, then they just started a baseline without checking for consistency (quadruple blinding, remember). So even if you had a previous EHF audiogram, it didn't have to match the one they would take on baseline.

The other possibility is the IHC theory.

So again only 3 possibilities (I think it's absolutely absurd to think placebo would allow you to hear literally twice as well) with regards to hearing improvements (word score improvements, sometimes doubling) all involving someone lying. Is it:

1) Frequency Therapeutics
2) Phase 1 super responders
3) Desperate Phase 2 applicants.

Depends on what you believe I guess but eventually it will be known.

 

[Zugzug]

I agree that 34% of patients responding with a 10% increase doesn't look great in numbers, but we know that even a 10% improvement translates to a huge QoL improvement. It also says nothing about the patients' QoL who didn't have a 10% increase but still experienced, subjectively speaking, some measure of benefit.

To me, and I'm just calling it straight, absolute WR improvement of >=10% is so little. It's a horrible metric, and way less convincing than the 95% CI (Thornton and Raffin). I am attaching some pictures with the 95% CI under the Binomial Distribution assumption. The first column represents baseline score and the second column is the range of the 95% CI.

As an example, if the baseline is 42%, they would need to exceed 60% to be statistically significant (positively). That's an absolute WR improvement of 18%.

If we look at the chart, the only time the upper bound is more than the baseline by <10% (in other words, the >=10% absolute improvement is more rigorous than Thornton and Raffin) is baseline percentages of 0, 2, and 88%+. But these scores likely weren't a part of the trial anyways.

With this being said, I still stand by my claims that exceeding the 95% CI on Thornton and Raffin (4 patients in Phase 1) is difficult to do -- particularly the 3 patients that crushed the upper bound. No way this is just from chance. Either cheating or FX-322 helping them, for sure.

I'm putting a lot more weight on the double blind Severe Hearing Loss Phase 1b trial than I am on the categorical, unblinded open-label study though. This study should help answer both the word score cheating issue and the "don't step on the lawn" issue.

 

[Ed209]

I am afraid you can't even fake that:

This is where fake WR scores were discussed after the phase I/II results. It's on page #161, so you can read it there. It's very interesting to see what is said compared to now.

The video that was posted, teaching you how to fake it, is in the quoted post (I haven't watched it, so I'm not sure what is said).

It is disappointing that they weren't more rigorous in this regard to stop this being a possibility in Phase 2a, especially with (retail) investors' money being at stake. They wouldn't care about that, though, as long as the more pivotal investment partners got their money out safely first before the drop occurred. There's no definitive proof of this, but something does stink a little bit. Forget the money, they should have nailed this trial design for their own scientific integrity.

 

[Zugzug]

@Zugzug, I believe this is not necessarily the case, but someone please correct me if I'm wrong. My understanding is that there was an initial screening test for all patients to determine whether you could participate in the trial or not. Once you had been screened (and selected), you then undertook the same tests again. These second tests were what were used as baseline - not the screening tests. What I find mind boggling is that in one scenario, these patients would have had to have lied twice: once during their screening and again during baseline. What I'm thinking though is, how likely is that given that once you know you've been admitted to the trial it doesn't matter how you perform on the new baseline test? What it really comes down to then is whether Frequency Therapeutics were aware of either scenario. The thing is, they weren't - they were blinded. It's not as if Frequency Therapeutics would have caught on to higher baselines that were different from the screening tests. So I can see one of the two possibilities here. Either patients lied/faked their results twice OR they faked their initial screening and then performed as normal, but by the time they performed their baseline Frequency Therapeutics were already blinded so it was out of their control. In the latter scenario, it's possible that patients who were originally thought to have word minimum WR deficits actually were way too close to the ceiling. In either scenario, Frequency Therapeutics would have only been able to figure out what was going on by cross-referencing against historical records. To put it simply, either scenario is a shitshow if I've ever seen one

So the answer to this question has a simple game theory solution. Keep in mind though, human beings are not smart and many don't use perfect rational decision making.

Say I made it into the trial. It doesn't matter how; I'm in. There's a 75% chance I will be treated and a 25% chance I won't.

If I reasoned to myself that "I'm going to tank my baselines," there are two reasons why I might be thinking this -- both are stupid. One might be to stay consistent with low screening and the other might be "there's a 75% chance that I'm in the FX-322 group. If I perform low at baseline and then crush the tests with treatment, I have a chance of helping the drug clear the finish line."

For the first one, there's no real motivation to stay consistent; one can't prove it or punish them. For the second one, it doesn't make sense because if they believed they were in the treatment group, they wouldn't have the same motivation for getting the drug through because they thought they were already receiving it.

If they then asked themselves "what if I'm in the placebo group?" then they wouldn't want to inflate placebo improvements, reducing the chances of eventually receiving the actual drug someday.

With all of this said, if someone cheated, I would think the extent of it would be exaggerating to get in and then performing the trial normally. This is immoral (and backfired), but it's at least a smarter form of cheating.

This would definitely appear to be part of the reason as to what went wrong. Coming back to your example, what if we imagined the complete opposite: that all or most patients lied/faked their results? In this case, the distribution would almost seem to not matter at all. Every patient in every group is unreliable, regardless of what the end result is. The whole trial would be worthless and the only way around it is to start again.

I basically answered this idea to @FGG, and I agree with you that if the widespread cheating was way over the top, you would be right. I don't think it was that much.

 

[andreimatei]

Don't want to be mean, but B. Riley's $35 revision was prior to the Phase 2a results.

Correct me if I'm wrong, though.

I'm seeing it as of this morning...

https://www.streetinsider.com/dr/news.php?id=18169286

 

[HootOwl]

What I find mind boggling is that in one scenario, these patients would have had to have lied twice: once during their screening and again during baseline. What I'm thinking though is, how likely is that given that once you know you've been admitted to the trial it doesn't matter how you perform on the new baseline test?

Interesting question... I suppose it would depend on how afraid you would be about getting "caught" should any discrepancies surface and how aware you are of the blinding process.

If you're unfamiliar with the extent of blinding, you might think that the trialing center (or the researchers) would be comparing your initial audiogram - that permitted you admittance to the trial - to the second audiogram created as a baseline for the trial itself.

I'm kind of turning this over in my head now.

The other scenario would be as you described. But - if they only lied once on the initial assessment, and not at the second baseline, once the data is unblinded Frequency should be able to see that clearly. Maybe what you proposed is true, and Frequency was blinded as soon as the baseline audiogram was created so they were never able to catch any fraudulent activity.

 

[Christine2222]

When are the results expected for the double blind Severe Hearing Loss Phase 1b trial?

 

[Toby1972]

There is too much speculation here again, regarding the results perhaps not being that bad. Oh no.

In addition to the rather confusing and inexplicable results (placebos), FX-322 simply did not show any effect that could somehow help us with tinnitus. Period.

10% better WR is nothing. Such small improvements can simply be due to the daily fluctuations. Or equipment. It's just nothing.

Worst of all, there were 0.0 improvements in the audiogram, which was my last remaining hope for tinnitus improvement.

Some FREQUENCY THERAPEUTICS people are now millionaires because of their stock sales before March 23rd. Probably mostly from the money of people who were emotionally invested. That hurts even more.

Maybe because of my warnings, especially on March 22nd, I saved a few people here from great losses. That's the only positive thing for me.

Due to my hopelessness, I unfortunately had the first glimpses of mean depression today, which I had got rid of with hard work for half a year.

I can't hide anything from my girlfriend either. We're already talking openly about my handicap, but of course I downplayed the whole thing. "It just annoys me sometimes", but "it doesn't affect me", but unfortunately it does way too much. For example, I haven't slept through for a mostly a year. I'm always tired. Before that I could sleep like a koala. Any music, and I love music, sounds terrible with a scratchy ear.

I always lived in the hope that somehow relief could be found.

I would love to quit my job and spend the whole day in front of my computer with masking sounds and beer to be distracted.

But I can't give up on my girl. Not this girl.

Unfortunately, my sarcasm doesn't help me anymore.

It's a shame I can't invite you to Oktoberfest. Maybe someone will invite me to the USA at your expense. After all, I am already vaccinated with COVID-19, as a high-risk patient (I only have 2 healthy lung lobes). Well, I still have a bit of sarcasm in me.

So, I'm starting to get really drunk today. Better than ADs.

 

[Tezcatlipoca]

Hasn't every "single dose" trial been successful?

 

[FGG]

So the answer to this question has a simple game theory solution. Keep in mind though, human beings are not smart and many don't use perfect rational decision making.

Say I made it into the trial. It doesn't matter how; I'm in. There's a 75% chance I will be treated and a 25% chance I won't.

If I reasoned to myself that "I'm going to tank my baselines," there are two reasons why I might be thinking this -- both are stupid. One might be to stay consistent with low screening and the other might be "there's a 75% chance that I'm in the FX-322 group. If I perform low at baseline and then crush the tests with treatment, I have a chance of helping the drug clear the finish line."

For the first one, there's no real motivation to stay consistent; one can't prove it or punish them. For the second one, it doesn't make sense because if they believed they were in the treatment group, they wouldn't have the same motivation for getting the drug through because they thought they were already receiving it.

If they then asked themselves "what if I'm in the placebo group?" then they wouldn't want to inflate placebo improvements, reducing the chances of eventually receiving the actual drug someday.

With all of this said, if someone cheated, I would think the extent of it would be exaggerating to get in and then performing the trial normally. This is immoral (and backfired), but it's at least a smarter form of cheating.

I basically answered this idea to @FGG, and I agree with you that if the widespread cheating was way over the top, you would be right. I don't think it was that much.

Word scores and audiograms don't have to match once you are in. It's quadruple blinded and they don't kick you out once you are in. You just have to pass the initial screen and maybe that's the problem (this should again be really obvious once unblinded individual data is out)...

Also, I stand by my comment that when you are looking for non severe audiograms but also want at least moderately bad word scores, you are selecting for a relatively narrow population of genuine patients. This puts the odds of "selecting for liars" at much higher than it would otherwise be. Not to mention Frequency Therapeutics messing up by telling people word scores would be an important part of Phase 2a criteria to avoid the ceiling effect seen in some of Phase 1 participants.

I agree that the severe trial will be illuminating.

In the meantime, there is no scenario where one of the 3 groups isn't lying unless you believe the placebo effect can double word scores (I don't).

 

[Zugzug]

Nothing has changed.

People mustn't get too emotionally invested in this. See it for what it is until there's some clinical evidence to show that it works. Until then, treat it the same way as any other drug that's in development, and be brutally objective.

I really feel for all of you that are depressed by the results, but that's biotech for you. Expect the unexpected in either direction.

I wasn't impressed with the phase I/II results, but it was primarily a safety trial, so I didn't really read too much into it in terms of efficacy. I noticed the issue was raised about people faking their results, but there was talk of the WR scores being falsified back in the Phase I/II trial. I remember someone posting a video on how to "fake it." The 10 dB improvements that were seen in 4 people had no clinical value. I believe there was an increase seen at 8 kHz; my audiogram had the same increase at the same frequency and I did nothing other than take a few vitamins. This is when I first started to doubt how effective FX-322 might be, but it was only a small cohort. Before those results were published, everyone was hyper-bullish saying more doses will equate to better results, etc, etc. I thought people at the time were getting too far ahead of themselves. I also didn't like how they spun the success of those earlier results to the media before they showed any data. The headlines and sound bytes had everybody salivating.

The latest results seem to be more of the same. They should have doubled down on the trial design and made sure the WR scores couldn't be faked. What I ultimately wanted to see was the slightest hint of an improvement on the audiograms. I realise audiometry is a tad antiquated, and that there could well be an improvement in hidden hearing loss, but I think people generally expected far more than that; especially in the earlier days.

Sorry for sounding negative, because it may come across that way, but those are my honest thoughts and feelings.

I think we all agree, from any angle, something odd has to be going on to double word scores. Whether it be cheating, "learning the test," or taking a drug (FX-322 possibly) that helps. A person (with stable hearing loss) doesn't just honestly double their score by chance.

This is all so depressing, honestly, because it's such a challenging problem to fix. All of the fixes that I can think of drastically lower the power of all statistical tests. In other words, they could make their design really strong with low chances of inconsistencies from screening to baseline, but the bar for FX-322 will then be too high for what's ultimately an adjunctive treatment for hearing loss. I'm sure the company is frustrated by this as well.

 

[Aaron91]

So the answer to this question has a simple game theory solution. Keep in mind though, human beings are not smart and many don't use perfect rational decision making.

Say I made it into the trial. It doesn't matter how; I'm in. There's a 75% chance I will be treated and a 25% chance I won't.

If I reasoned to myself that "I'm going to tank my baselines," there are two reasons why I might be thinking this -- both are stupid. One might be to stay consistent with low screening and the other might be "there's a 75% chance that I'm in the FX-322 group. If I perform low at baseline and then crush the tests with treatment, I have a chance of helping the drug clear the finish line."

For the first one, there's no real motivation to stay consistent; one can't prove it or punish them. For the second one, it doesn't make sense because if they believed they were in the treatment group, they wouldn't have the same motivation for getting the drug through because they thought they were already receiving it.

If they then asked themselves "what if I'm in the placebo group?" then they wouldn't want to inflate placebo improvements, reducing the chances of eventually receiving the actual drug someday.

With all of this said, if someone cheated, I would think the extent of it would be exaggerating to get in and then performing the trial normally. This is immoral (and backfired), but it's at least a smarter form of cheating.

I basically answered this idea to @FGG, and I agree with you that if the widespread cheating was way over the top, you would be right. I don't think it was that much.

Great breakdown, you articulated it very well. To your point though, this would be assuming that entrants are acting rationally. I guess we can never know what they were thinking.

I think then the only possibility is that they tanked the initial screening test to get in and performed as normal once they were in. As you said, there was no incentive to tank the test again. Once you marry this idea with Frequency Therapeutics saying they checked historical records after the shocking results, you begin to realise that this probably did happen. If Frequency Therapeutics weren't blinded, surely they would have caught on to the discrepancy between the screening test and the baseline test. So what they must have found (after unblinding) was this: the baselines matched historical records, but not the screening tests, meaning that they essentially had participants in the trial who did not meet the eligibility criteria. I think it is fair to say that once you have a sample population that doesn't fit the criteria you are trying to control for, the whole trial is worthless. I think I'm coming round to the idea that Frequency Therapeutics were not lying.

As for the extent of cheating, I guess this remains a question mark, but for Frequency Therapeutics to have used it as an excuse in the first place, it would have to be, by definition, widespread. It was telling to me that Lucchino said they had combed through social media and found people sharing info about what it took to get into the trial. Make no mistake: Frequency Therapeutics have gone through this thread over the last month since they got the results. There's no doubt in my mind about this. FGG is right in that someone (be it Frequency Therapeutics, Phase 1 patients or Phase 2 patients) is lying, but I think the only question that needs answering here is whether we believe Frequency Therapeutics are credible. There was clearly no incentive for Phase 1b patients to lie, so that trial was legit as far as I'm concerned. It's then very binary. If you infer Frequency Therapeutics are credible, the answer is clear: Phase 2 patients lied. Given Frequency Therapeutics' pedigree, I'm inclined to say they are credible, despite this awful readout.

Edit: @Zugzug, @HootOwl has just made a very good point about your point on motivation to stay consistent between the two tests. It's possible the patients didn't know whether they were being "checked" for consistency, even though the trial was clearly stated to be blinded at all levels. I think it's possible some patients just wouldn't know/wouldn't care what blinding means, but maybe I'm wrong. Either way, it doesn't matter. The alternative scenario I also described still screws things up as well.

 

[Olive27]

The craziest part of all of this is that someone definitely lied because people don't double word scores from placebo.

So it depends on which of the 3 you think did:

1) Frequency Therapeutics themselves
2) Phase 1 participants
3) Phase 2 participants.

Frequency Therapeutics says they have direct evidence it is option 3, which would mean the drug does work. Time will uncover this.

But the drug doesn't work... hence no improvements in the EHF audiogram in the recent FX-322-111 trial. It's all placebo unfortunately.