@Diesel has already pretty much answered this above but I thought I'd just expand a little bit more on this for what it's worth as I've seen this question come up a lot of times from those who consider themselves more in the "mild" group of hearing loss cases and have concerns as to whether they would benefit from FX-322.
There is a key differentiation here to be made between
statistically significant improvement and what one perceives in their own experience as significant improvement. One of the key principles of proving something in science
with confidence (or strictly speaking,
disproving something with confidence) is to show that an outcome cannot be explained by chance, or at least cannot be explained by chance most or all of the time.
With this in mind, if one were to take a word-in-noise test several times, one would not always necessarily achieve the same score. As an arbitrary example, one with mild hearing loss may take a word-in-noise test several times and on the first occasion they may score 45/50. However, on on the second and third try they may score 42/50 and 48/50 respectively. In other words, scores will vary - perhaps even improve - even in the absence of a treatment. This test variability is what is known in statistics as a "standard deviation". So for one to state confidently that an improvement in a word score is down to treatment and not chance, one would need to see an improvement of
at least more than one standard deviation, depending on where one draws the line between chance and real improvement (and this line is usually guided by certain mathematical standards).
Now, in a x/50 word-in-noise test, Frequency Therapeutics considers 3/50 words to be the measurement of one standard deviation. Depending on how confident one wants to be, one may require to see two, three or more standard deviations post-treatment to be able to say "we can say with X amount of confidence that this improvement cannot be explained by chance". Now, I don't remember where Frequency Therapeutics' confidence level was drawn or whether there is a set confidence level for clinical trials, but assuming it was at say 2 standard deviations (6 words), and a patient went from a 45/50 word score to 50/50 (5 words), Frequency Therapeutics would not be able to say they were confident enough that the improvement was not down to chance because 5 words is less than 6 words, and therefore less than two standard deviations. In other words, the "ceiling" for seeing
statistically significant improvement in mild hearing loss cases in the phase 1 study was too low, even though there was/may have been (don't remember now) evidence of actual word score improvement.
This is why Frequency Therapeutics are focusing more on the moderate to severe hearing loss groups because they know the ceiling for improvement is much higher in these cohorts and they are likely to achieve much more favourable, statistically significant results - and it's statistical significance that counts when it comes to proving a drug is worthy of going to market.
TL;DR: mild-hearing loss cases should still benefit from FX-322, even if clinical trial results do not show improvement at a clinically/statistically significant level.