Great post. I was also noodling on the idea of a central entity coordinating and running trials for multiple treatments that target the same disease, but came to the same conclusion as you that the current throughput isn't high enough and the time lag between development reduces the chances of a shared control arm. I also believe that the AI biotech push will rapidly increase the number of plausible candidates per year to the point where this does become feasible.
Good call out on the third wave of philanthropic funding and its implications! Something I didn't think about. Will also throw focused research organizations (https://www.convergentresearch.org/) into the mix as a potential funder that I've been thinking about
Yes, FROs do belong on the list here. The time lag/concurrency issue will vary a lot case-by-case so I didn't want to get into it in this blog, but several people asked about it, so I think we will put out a separate blog post on it sometime!
The checkpoint inhibitor comparison is the part that sticks, because it shows the cost of not combining market and trial design even when the science succeeds: a wave of similar drugs, each with its own hugely expensive trial fine-tuned to show a narrow advantage, and clinicians still left without a clear answer on sequencing or combination.
That's the strongest case for a shared platform, since it isn't only about efficiency for the funder, it's about producing the head-to-head answer fragmented trials structurally can't generate.
I think a lot of public funding subsequently went into learning how to actually use these treatments, so one wonders how you could have had a smoother transition between private and public evidence generation.
Maybe the smoother path is not treating approval as the finish line for the shared design. If the AMC funder is already running the platform trial pre-approval, that infrastructure and the relationships with participating firms already exist. Extending the same platform into a pragmatic, adaptive phase (dosing, sequencing, combinations) as a built-in second stage of the original commitment, rather than a separate public program assembled after the fact, would mean the "how do we use this" evidence gets generated on the same rails as the "does this work" evidence, instead of public funders reconstructing that infrastructure from scratch once approval hits.
Great comment. Yes, that's the ideal, as far as I am concerned. I've tried pointing to this in a couple of places in this note. But it's also a pretty distant prospect so for now I didn't want to put too much weight on it.
That vaccine case sharpens it more than the checkpoint one, since it's the same failure in a field we're already good at: the bottleneck isn't capability, it's the handoff. If optimisation takes decades because it's a separate step assembled after approval, the decades are the price of that handoff, not of the problem being hard.
You are an expert on clinical trials and I would like you to take a look at the results of the Emory University FDA montelukast Alzheimer's clinical trial posted on clinicaltrials.gov. The trial was set up so that the treated group was to take 20 mg twice a day of the anti-inflammatory leukotriene inhibiting asthma drug montelukast for one year. The final results submitted in 2024 indicated that the placebo group showed cognitive improvement after one year and the montelukast treated group showed cognitive decline. My contention is that the trial was sabotaged by switching the montelukast and placebo dosages between the groups before the start of the trial. Emory University did not publish the results for obvious reasons. I notified the FDA Office of Scientific Investigations in January 2024 but they do not appear to have investigated it. Please take a look.
The NIH Toolbox Cognition Battery shows the montelukast treated group went from 35.4 to 33.0, which indicates a cognitive decline. The placebo group went from 35.4 to 37.3 and showed improvement, not something you would expect to see. Further, cerebrospinal fluid (CSF) amyloid and tau levels declined in the placebo group, positive indicators that were also not expected. There were no serious adverse events in either group.
By way of comparison, Biogen's FDA approved drug lecanemab (Leqembi) never showed cognitive improvement, but only showed a slight slowing of decline. The placebo group in the Emory trial showed cognitive improvement which indicates that the montelukast dosages were given to the placebo group.
Another clinical trial in Canada using 30 mg montelukast twice a day showed cognitive improvement in its preliminary results but was soon forced into bankruptcy by its largest creditor Atai Life Sciences before the completion of the final results.
In a small clinical study published in April 2025, fifteen Parkinson's disease patients were recruited to the Karolinska Hospital in Sweden for a 12-week open-label trial of 20 mg twice a day montelukast treatment. .
All patients completed the study. No serious adverse events related to treatment were reported. MDS-UPDRS total scores improved by 6.9 points. The Montreal Cognitive Assessment (MoCA), a test often used to assess Alzheimer's patients' cognitive abilities, showed significant cognitive improvement.
The Karolinska Montelukast Parkinsons Clinical Trial in Sweden began in February 2024 and the estimated end of the trial is 31 December 2026. The purpose of the trial is to see if high dosages of montelukast given twice a day slow down Parkinson's disease progression. Participants in the treated group are receiving up to 30 mg of montelukast twice a day in the form of patented VersaFilm buccal tablets developed by Intelgenx Technologies and now owned by Atai. Participants consist of up to 90 people with early to moderate PD who are also on levodopa. The duration of treatment is 18 months, followed by a 3 month washout.
Atai halted Intelgenx's Alzheimer's clinical trial before the final results were concluded. Atai is now supplying Intelgenx's version of montelukast the Karolinska Parkinson's clinical trial. I am concerned that Atai's version of montelukast does not contain the correct amount of montelukast as stated on the packaging.
I have been taking 10 mg montelukast multiple times a day for extreme mental fatigue since 2016. I got completely back to normal within one week after I started this treatment. I am presently taking 20 mg three times a day. I am a retired American living in Ghana where I can buy as much montelukast as I need without a prescription. My mother passed away with Alzheimer's and I believe that my taking montelukast multiple times a day has prevented me from going down that path. My website is www.montelukast-repurposed.org. I have more evidence on my site about who may have been envolved in the sabotage.
Great post. I was also noodling on the idea of a central entity coordinating and running trials for multiple treatments that target the same disease, but came to the same conclusion as you that the current throughput isn't high enough and the time lag between development reduces the chances of a shared control arm. I also believe that the AI biotech push will rapidly increase the number of plausible candidates per year to the point where this does become feasible.
Good call out on the third wave of philanthropic funding and its implications! Something I didn't think about. Will also throw focused research organizations (https://www.convergentresearch.org/) into the mix as a potential funder that I've been thinking about
Yes, FROs do belong on the list here. The time lag/concurrency issue will vary a lot case-by-case so I didn't want to get into it in this blog, but several people asked about it, so I think we will put out a separate blog post on it sometime!
Put me on the list of people who'd be very interested in reading about it!
The checkpoint inhibitor comparison is the part that sticks, because it shows the cost of not combining market and trial design even when the science succeeds: a wave of similar drugs, each with its own hugely expensive trial fine-tuned to show a narrow advantage, and clinicians still left without a clear answer on sequencing or combination.
That's the strongest case for a shared platform, since it isn't only about efficiency for the funder, it's about producing the head-to-head answer fragmented trials structurally can't generate.
I think a lot of public funding subsequently went into learning how to actually use these treatments, so one wonders how you could have had a smoother transition between private and public evidence generation.
Maybe the smoother path is not treating approval as the finish line for the shared design. If the AMC funder is already running the platform trial pre-approval, that infrastructure and the relationships with participating firms already exist. Extending the same platform into a pragmatic, adaptive phase (dosing, sequencing, combinations) as a built-in second stage of the original commitment, rather than a separate public program assembled after the fact, would mean the "how do we use this" evidence gets generated on the same rails as the "does this work" evidence, instead of public funders reconstructing that infrastructure from scratch once approval hits.
Great comment. Yes, that's the ideal, as far as I am concerned. I've tried pointing to this in a couple of places in this note. But it's also a pretty distant prospect so for now I didn't want to put too much weight on it.
I wrote about it in the past in the context of optimising vaccines: https://www.cgdev.org/publication/more-less-optimising-vaccines-constrained-world We are good at doing it, but it can take decades, because the approval and optimisation steps are completely separate
That vaccine case sharpens it more than the checkpoint one, since it's the same failure in a field we're already good at: the bottleneck isn't capability, it's the handoff. If optimisation takes decades because it's a separate step assembled after approval, the decades are the price of that handoff, not of the problem being hard.
You are an expert on clinical trials and I would like you to take a look at the results of the Emory University FDA montelukast Alzheimer's clinical trial posted on clinicaltrials.gov. The trial was set up so that the treated group was to take 20 mg twice a day of the anti-inflammatory leukotriene inhibiting asthma drug montelukast for one year. The final results submitted in 2024 indicated that the placebo group showed cognitive improvement after one year and the montelukast treated group showed cognitive decline. My contention is that the trial was sabotaged by switching the montelukast and placebo dosages between the groups before the start of the trial. Emory University did not publish the results for obvious reasons. I notified the FDA Office of Scientific Investigations in January 2024 but they do not appear to have investigated it. Please take a look.
Emory clinical trial
https://clinicaltrials.gov/ct2/show/NCT03991988
The NIH Toolbox Cognition Battery shows the montelukast treated group went from 35.4 to 33.0, which indicates a cognitive decline. The placebo group went from 35.4 to 37.3 and showed improvement, not something you would expect to see. Further, cerebrospinal fluid (CSF) amyloid and tau levels declined in the placebo group, positive indicators that were also not expected. There were no serious adverse events in either group.
By way of comparison, Biogen's FDA approved drug lecanemab (Leqembi) never showed cognitive improvement, but only showed a slight slowing of decline. The placebo group in the Emory trial showed cognitive improvement which indicates that the montelukast dosages were given to the placebo group.
Another clinical trial in Canada using 30 mg montelukast twice a day showed cognitive improvement in its preliminary results but was soon forced into bankruptcy by its largest creditor Atai Life Sciences before the completion of the final results.
https://www.clinicaltrialsarena.com/news/intelgenx-alzheimers-disease-trial/
In a small clinical study published in April 2025, fifteen Parkinson's disease patients were recruited to the Karolinska Hospital in Sweden for a 12-week open-label trial of 20 mg twice a day montelukast treatment. .
All patients completed the study. No serious adverse events related to treatment were reported. MDS-UPDRS total scores improved by 6.9 points. The Montreal Cognitive Assessment (MoCA), a test often used to assess Alzheimer's patients' cognitive abilities, showed significant cognitive improvement.
https://pmc.ncbi.nlm.nih.gov/articles/PMC12006882/
The Karolinska Montelukast Parkinsons Clinical Trial in Sweden began in February 2024 and the estimated end of the trial is 31 December 2026. The purpose of the trial is to see if high dosages of montelukast given twice a day slow down Parkinson's disease progression. Participants in the treated group are receiving up to 30 mg of montelukast twice a day in the form of patented VersaFilm buccal tablets developed by Intelgenx Technologies and now owned by Atai. Participants consist of up to 90 people with early to moderate PD who are also on levodopa. The duration of treatment is 18 months, followed by a 3 month washout.
Atai halted Intelgenx's Alzheimer's clinical trial before the final results were concluded. Atai is now supplying Intelgenx's version of montelukast the Karolinska Parkinson's clinical trial. I am concerned that Atai's version of montelukast does not contain the correct amount of montelukast as stated on the packaging.
https://parkinsonsnewstoday.com/news/phase-2-trial-montelukast-oral-film-starts-dosing/
https://euclinicaltrials.eu/ctis-public/view/2023-504278-39-00
I found out about montelukast as a possible treatment for Alzheimer's in 2016 from this "The Guardian" article.
https://www.theguardian.com/science/2015/oct/27/asthma-drug-could-rejuvenate-ageing-brains-study-suggests
I have been taking 10 mg montelukast multiple times a day for extreme mental fatigue since 2016. I got completely back to normal within one week after I started this treatment. I am presently taking 20 mg three times a day. I am a retired American living in Ghana where I can buy as much montelukast as I need without a prescription. My mother passed away with Alzheimer's and I believe that my taking montelukast multiple times a day has prevented me from going down that path. My website is www.montelukast-repurposed.org. I have more evidence on my site about who may have been envolved in the sabotage.