I am today speaking regarding this new treatment recently reported, the AMT-130. This is genetic therapy to reduce the concentration of mutant Huntingtin, which is believed to be responsible for the generation of Huntington’s disease, or at least that is the hypothesis. The results were presented recently. They were submitted to the FDA, and the FDA has issued a non-approval of the current way in which the data is presented...
I am today speaking regarding this new treatment recently reported, the AMT-130. This is genetic therapy to reduce the concentration of mutant Huntingtin, which is believed to be responsible for the generation of Huntington’s disease, or at least that is the hypothesis. The results were presented recently. They were submitted to the FDA, and the FDA has issued a non-approval of the current way in which the data is presented. And here I’ll present why it is the case that the FDA, in fact, had the right call with regards to what the data shows. So the first thing is to try to tell you that 75% is really quite good. If any therapy lowers the likelihood of progression by 75% or slows progression by that much, we should all be on it. So the point I want to try to make in reviewing the data is to show you that 75% is too good to be true because it isn’t true.
So let’s first review the scale. The scale is a 0 to 25 scale. The higher you go, the worse. Should be higher, worse. So 25 is your worst possible score. The observed range in most patients is 2 to 18. Patients in this particular trial were around 15. That’s not uncommon for trials. And what we need to know is that the mean annual change in this scale is a decline of about a point per year. And it’s been somehow suggested that a clinically meaningful change is a reduction of about 0.2 to 0.3 per year. This is from a reference in 2017 that calculated this based on a clinical trial.
So the first thing to know is that when you have a Phase II, you’re all about safety and typically is double blind, but this was open-label. And in fact, there was no comparison. So that group that you see here on gray wasn’t really there to begin with. This was people that were given in an open-label fashion and a brain surgery and then underwent a number of lumbar punctures. And as you can imagine, there is a lot that goes into it and a lot of expectations for it. So open-label is not the best way to get at what the efficacy of an intervention is. It is good to know if we’re using a safe medication, but not really good to make judgments on efficacy.
The second thing is that when we say 75% treatment effect, we are meaning relative risk reduction. So we’re using this comparator, which is a non-placebo cohort, it’s an observational cohort, and making inferences from them as to what the relative change based on that benchmark it is. And so if you look at the data from that perspective, you can say 75%. But in reality, what patients care is not what it is relative to another group, but it’s relative to themselves. And the relative difference from baseline to the end is quite limited. As you can see, it’s 0.38 at the end of three years. So 0.4 points at three years comes to about 0.13 per year, which, as we mentioned, is below what’s clinically relevant. So 0.4 change per three years, that’s 0.13, well below a meaningful change. And that’s very important. Again, that’s just accounting for the specific curve that matters to patients, where they were at the beginning of the trial, where they end at the end of the trial, as opposed to where the observational cohort that is non-placebo controlled would be, right? And the reason you want to have a concurved placebo is because the placebo effect is very high. So very, very important.
Now, the other thing to remember is that there have been only a limited number of patients exposed. In fact, over the 36 months, there were 12 subjects completed in the high-dose group. We don’t have data on the low-dose group. And so we have only 12 subjects at the end. And we also don’t have a sham comparator, which was the case for the first year. And we have no reports from what that sham comparator was. And this is important because, as you can imagine, if you have surgery, you will have very clear placebo effects. Any intervention is going to lead to placebo effects. That’s why stem cell therapies were really good on the open-label phase of their development. But once they were compared to sham stem cells, then, of course, we did not see that, at least in the prior iterations of stem cell research.
And then lastly, there isn’t really much data on the mechanism of action. We believe that the medication works by reducing the mutant Huntingtin levels. It may well be that that didn’t happen, or, as it was suggested in a prior research, that this might in fact have increased the levels, which potentially is something of interest to learn, but we don’t have data. So we don’t know if this trial showed the appropriate target engagement so that then we can make sense of the clinical changes that have occurred.
So this 75% slower decline is too good to be true because it isn’t true. Placebo is always an important effect in invasive open-label treatments. That’s why most open-label trials always seem to work. 75% is a relative change, not an absolute change, which is what matters to patients relative to a non-placebo observational cohort. That change was not clinically meaningful, as we’ve seen over a three-year period, really only 0.4 points. So that’s not quite where we would want this to be or any therapy to be. There is truly a very small sample. There is no sham cohort and no data on mutant Huntingin levels.
So for all these reasons, it is very clear that the medication wasn’t approved by the FDA because it wasn’t quite showing what it is potentially going to show. This is an early state of safety, not of efficacy. And the efficacy we have is only an indirect measure. And so far, not quite as robust as the FDA would want it. So this is a good decision. And this is the evidence behind that decision. Thank you.
This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.
