AD/PD 2026 | Phase II trial insights on ambroxol in Parkinson’s disease dementia

So the original trial was a double-blind placebo-controlled trial where we screened 75 patients with Parkinson’s disease dementia and randomized them to either placebo or 1,050 milligrams of ambroxol per day. We chose that dose because that was a dose that was safe in pregnancy. That’s lower than what many of the international studies are using now, but this was the safety data that we had. So we screened 75 patients, randomized 55. We found that ambroxol was pretty safe overall. There were some GI side effects, which we expected. And unfortunately, so our primary outcome was something called the ADAS-COG, which is quite a complicated cognitive scale. And what we found is the patients were very heterogeneous. They are all completely different entering the trial. And we had an enormous placebo effect with some patients on placebo gaining 10 points on an ADAS-COG, and that’s just something that never happens in the clinic. So overall, we didn’t see a benefit in the ADAS-COG. There were a small number of patients with GBA1 mutations, and they appeared to do better. Now, again, you have to be, this is only like four patients, so you have to be very careful drawing conclusions from that. But that was encouraging. When we looked at the neuropsychiatric inventory, which is a combination psychiatric scale rating, all kinds of symptoms, including psychosis and mood, what we found was the placebo patients got clearly worse in the initial trial. And there was a difference in the NPI between treated and control of almost six points, which is a huge clinically significant change. So that just missed statistical significance, but it looks like a really encouraging response. And we had approval in a plasma biomarker called GFAP, which again appeared better in the treated patients.

So in our open-label extension, which is what we’re presenting here, 34 patients who ended the double-blind randomized control trial were enrolled to get ambroxol, 1,050 milligrams a day for six months. So 17 placebo to ambroxol patients and 17 ambroxol to ambroxol patients. We had four serious adverse events, which we do not believe are related to ambroxol. We quickly achieved blood levels, which should have been therapeutic, just like the originally treated group in the double blind. And we also saw nice rises in the glucocerebrosidase enzyme, which we were following in white blood cells. Overall, the safety profile is similar. So we did see a little bit of stomach upset in the people who are changing from placebo to ambroxol, but this was mostly mild. The main findings of the trial were that the main trend lines that we observed in the double blind basically carried forward. So I don’t know if that means that six months isn’t enough to make a change. So again, no difference on the ADAS-COG, which is our primary outcome measure. On the NPI, the people who were on placebo and change to ambroxol, who were basically worse at the end of the trial, they got a little bit better or they trended in a line which followed the ambroxol to ambroxol group. And that group, again, just continued in a straight line from where they were doing. Again, so the patients with GBA1 mutations, which is, again, a very small number, did better on the ADAS-COG and the neuropsychiatric inventory. So again, this is encouraging. There’s a very, very small number, so we have to be careful with what we conclude. So where does this go? So my primary interest is in cognition, and we were trying to set up a trial in Lewy body disease. And during this interval, there are now large randomized controlled trials of ambroxol. And so there’s one in England by Dr Schapira, who’s looking at Parkinson’s disease, and he wants to enroll 330 patients. There’s another trial in Norway by Professor Arvid Rongve looking at Lewy body dementia. And so hopefully we’ll see results of that in the next few years. But we didn’t have definitive results where the treated patients were better, but we had a number of trends that looked good. And it’s a small number of patients in a very heterogeneous group of patients. So they were all very different going in. So we were actually sort of in trouble even before we did our analysis because everyone was so different.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.