>"The patients demonstrated improved brain function and showed a reduction in the protein plaque gumming up their neurons, the company’s report says."
Still chasing the amyloid hypothesis I see. This is a symptom, not a cause, so they are wasting their time (in terms of actually helping the patients, eventually someone will randomly get "positive" results and a bunch of money if enough trials are run).
> The patients demonstrated improved brain function
If this part of the quote is actually true, then what you are saying is either wrong or irrelevant. They appear to have evidence it is true, and are continuing to verify it.
There are various ways this could be helping (if it is):
1. The amyloids are a cause of Alzheimer's
2. The amyloids are a symptom of Alzheimer's, but are a cause of the memory impairments
3. The drug has a side-effect which helps the memory impairments.
In any case, what is your expertise? I love random comments that say that scientists who have spent decades in a field are simply "wasting their time" because they haven't read some article that was posed on the web. (NB, this may not be the case with you, if you are indeed an expert in the field.)
> "The patients demonstrated improved brain function"
So what? If you run trials with large enough sample size or allow sufficient researcher degrees of freedom (multiple ways of measuring "brain function", reporting results whenever they look best, looking at subsets of the patients, etc), then you have 50% chance of the treatment group "demonstrating improvement".
For example, from the limited information available I see in this case they measured "brain function" in 5 different ways:
Change from baseline in cognitive: Alzheimer´s Disease Assessment Scale- Cognitive (ADAS-Cog) [ Time Frame: 78 weeks ]
Change from baseline in function: Alzheimer´s Disease Cooperative Study-Activities of Daily Living (ADCS-ADL) [ Time Frame: 78 weeks ]
Change from baseline in cognition: Mini-Mental State Examination (MMSE) [ Time Frame: 78 weeks ]
Change from baseline in global assessment: Clinical Dementia Rating-Sum of Boxes (CDR-SB) [ Time Frame: 78 weeks ]
Change from baseline in behavioral assessment: Neuropsychiatric Inventory (NPI) [ Time Frame: 78 weeks ]
Not that I agree with this type of analysis but it is going to be similar to what they did to determine "demonstrated improvement":
Lets say there was a 2.5% percent chance of "randomly" getting a result like theirs. Then (1-0.025)^5 ~ 88% of the time there would be "no demonstrated improvement" and ~12% of the time there would be.
How many clinical trials for Alzheimer's like this have been done? We'd expect at least ~1/8 of them to "demonstrate improvement" just based on how the studies are designed.
>"In any case, what is your expertise?"
The reason it is a scholarly practice to cite your sources is that the audience doesn't need to rely on heuristics like this. Don't take my word for it, just follow the links and assess the published evidence for yourself.
It's a Phase II trial, which means it is focused on identifying dosage ranges and side effects in human rather than a large-scale efficacy trial. (It's also the first time you actually give a drug candidate to humans).
You also forgot that there is a fourth point: there is no actual effect, and the apparent effect is statistical fluke.
It's hard to understand what could be going on from this article alone, because it spends a whopping 1.5 paragraphs on explaining any details about the putative vaccine. The rest of the piece reads as a generic PR fluff piece, which makes me think that this was largely based on company-provided PR and so anything approaching details is based on the most positive light that they can be spun into.
> ... a Phase II trial, which means it is focused on identifying dosage ranges and side effects in human rather than a large-scale efficacy trial. (It's also the first time you actually give a drug candidate to humans).
FWIW your drug candidate is administered in all phases. Phase I you look for side effects ("safety and tolerability"). Phase II is, as you say, a dose ranging study, but you can't exceed dosages you've already tested for safety. You usually have multiple arms (a placebo and a couple of dosages); when you find an seemingly efficacious dose you then proceed to a Phase III where you test that dosage (or several) on a statistically representative enough sample to hopefully predict results in the population at large.
(former drug designer who has designed some Phase I and II studies).
==
Fun fact: when you file your IND with the FDA to test a new drug you aren't actually asking for permission. You're asking them to tell you that if you follow the protocol you have sent them, based on the evidence you've sent them from what you know already, they won't prosecute you for violating the law on administering unapproved drugs. Section 505 for the food and drug law has a bunch of exceptions -- you ask then to tell you you fit in one of the exceptions.
And if they don't tell you "no" in a 30 days...you can go ahead! I wonder if the shutdown extends that 30 days.
Thanks for the correction. I thought Phase I was a pre-human clinical trial for safety, rather than a human clinical trial. I appear to have been mistaken.
No problem. Preclinical work is indeed called preclinical work :-).
Typically you do a bunch of in vitro and in vivo studies when developing your drug candidate. When you're ready to file the IND (asking permission to test on humans) you do a different set of studies specifically to generate the data you need for your submission. Sometimes the agency will even tell you where its concerns may lie so you might use a specific animal or specific test to address them. You also include a lot of in vitro work about how you manufacture (and control the mfg of) your test product.
The difference might be that you take and process a lot of samples in your R&D stage, and then when you prepare your submission you get independent labs to do the work (to show the agency you don't have your "finger on the pan" so to speak). For example on a depot-injectable drug I worked on we did six experiment sites per guinea pig (because we knew tissue diffusion would be only a few millimeters) which saved us time and a lot of money but when preparing for the FDA we did only one per, so we could guarantee to them that any trace of drug found anywhere in the tissue came from a single injection. Likewise we did a lot of pathology ourselves just peering through the microscope (and learned a lot!), but for the FDA we had the animal lab send tissue samples directly to an external pathologist who then sent us a report without any input from us. Etc.
Honestly, by pointing to your comments from previous discussions, you look like you have an ax to grind rather than being helpful.
The truth is that we don't know what causes Alzheimer's. Amyloids may or may not play as much of a role as some people think. But without a better hypothesis to test and get funding for, this is what they can do.
Considering that every single anti-amyloid Alzheimer's candidate has failed--even those that made it into Phase III trials--anyone suggesting something that attacks amyloids has to overcome a lot of skepticism.
Assuming what you say is true, HN continues its trend of having great content for software and related topic, and being next to useless for meaningful physics or medical news. I didn't notice this till my buddy who is a physicist pointed out that almost all of the physics articles on here are either wrong, or wildly misleading. It would be nice if we could be better about vetting this stuff as a community.
Don't take my word for it. Follow the links to what the primary literature says.
>"being next to useless for meaningful physics or medical news"
Well, to be fair the "amyloid hypothesis" is still the mainstream hypothesis. It's just that this is for reasons that are baffling to me. Afaict, it has nothing going for it at all vs the prior "cholinergic" hypothesis (which has resulted in treatments that actually appear to do something for the patient).
> the "amyloid hypothesis" is still the mainstream hypothesis.
I agree with you. Amyloid may be the most popular but it isn't what it used to be 10-15 years ago (dogma). it has much weaker support. What is particular true and sad about what you've said is that it probably is the most popular but only due to lack of consensus on anything else. After running into the amyloid wall running at full-speed about a dozen times, significant chunks of pharma have nearly abandoned working on the disease, and put some funds into external research.
No, just that it isn't going to be removing amyloid plaques. For the plaques to build up the neurons are clearly malfunctioning for some reason. Beyond that, who knows?
I always forget about reader mode. I'll try to remember it more in the future. It let me read this site without having to worry about enabling javascript or dancing around ad-blocker issues.
Still chasing the amyloid hypothesis I see. This is a symptom, not a cause, so they are wasting their time (in terms of actually helping the patients, eventually someone will randomly get "positive" results and a bunch of money if enough trials are run).
1) Up to 30% of cognitively normal elderly have detectable amyloid deposits: https://news.ycombinator.com/item?id=17444214#17453147
2) Only 70% of dementia patients have detectable amyloid deposits: https://news.ycombinator.com/item?id=17444214#17445337
3) Amyloid-related genetics can only account for < 10% of Alzheimer's cases: https://news.ycombinator.com/item?id=17444214#17457198
4) Amyloids are the most thermodynamically favorable structure that polypeptides can form (ie require constant vigilance by the cell to prevent): https://news.ycombinator.com/item?id=14914528#14917057
5) Can anyone find a disease state where they looked for amyloids and didn't find a positive correlation? I doubt it.