4 ms·

The article is not quite updated. Biogen did end their phase 3 trial of the amyloid-beta antibody Aducanumab early, citing lack of efficacy. A closer look at th

by zosima 7y ago

The article is not quite updated. Biogen did end their phase 3 trial of the amyloid-beta antibody Aducanumab early, citing lack of efficacy. A closer look at the numbers however appeared to reveal that they shouldn't have. And Biogen is now filing Aducanumab for approval by FDA. [1] All in all, there is no question that amyloid-beta is a key part of Alzheimer. Already Alois Alzheimer, when first describing Alzheimer's disease, noted amyloid-beta plaques as a key pathological finding. Amyloid-beta plaques are a characteristic of the disease, and a proper Alzheimer diagnosis requires the presence of amyloid-beta plaques. Many variants of early-onset Alzheimer are caused by mutations either directly in APP (the precursor to amyloid-beta) or genes intimately involved with APP-processing (e.g. PSEN1 and PSEN2). These mutations often cause extreme levels of amyloid-beta and amyloid plaques. Raised levels of amyloid-beta and amyloid plaques is a necessary condition for Alzheimer's disease, both early- and late-onset. But it's clearly not a sufficient criteria: many people have brains full of plaques and amyloid-beta, but little or no cognitive decline. So the problem with Alzheimer's research is not so much the understanding that amyloid-beta is a significant part of the story, but rather focusing on it as the only part of the story and the only conceivable solution. [1] https://edition.cnn.com/2019/10/22/health/biogen-alzheimers-drug-fda/index.html https://edition.cnn.com/2019/10/22/health/biogen-alzheimers-...

amluto 7y ago
Nothing says p-hacking like a drug company canceling a trial and then, a few months later, re-analyzing the data and magically finding a benefit.
- zosima 7y ago
  FDA usually puts some quite strict limits on the amount of p-hacking and general shenanigans they allow in clinical trials. But there's no approval yet, and no data to look at, so it's too early to judge. Anyway, usually Alzheimer trials are a jumble of noise. The cognitive measurements are imperfect and patients are improving and deteriorating randomly, taking a very different disease course than you'd expect from textbooks. Finding a signal in the data, even for a working drug can be quite hard. I could just as well believe that Biogen ended the trial hastily because all competitors had failed and they didn't do a thorough analysis.
  - memling 7y ago
    Derek Lowe's comments on the reboot were understandably a bit skeptical: > [W]hile the new larger EMERGE data do look like there’s a dose-responsive effect in the primary measurement (the CDR-SB scale) and three secondary dementia scores, the ENGAGE data actually show worse than placebo for CDR-SB at the higher dose (see slide 19 in Biogen’s presentation). They’re also worse than placebo in the MMSE scoring, but both the ADAS-Cog13 and the ADAS-ADL-MCI scales show apparent dose-responsive improvement. I have no idea why four different dementia rating systems should give such different results, and to a first approximation, no one else seems to have a good idea, either. So we’ve got the EMERGE/ENGAGE difference to explain (you’ve seen Biogen’s explanation, higher dosing later in the trial), and the variable results within the ENGAGE trial itself. And who are these high-dose patients in the ENGAGE study that Biogen says line up with the EMERGE trial? Ed Silverman at Stat quotes an analyst saying that it looks like fewer than 100 patients out of the whole set. That reminds me a bit of Biogen and Eisai’s tease of “We’re seeing positive data!” during the clinical trials, which also turned out to be based on a smaller number of patients than one would want when making a call on Alzheimer’s efficacy.[1] He's written a fair amount about Alzheimer's,[2] and it's been an informative and useful commentary (at least for me). [1] https://blogs.sciencemag.org/pipeline/archives/2019/10/23/the-return-of-aducanumab https://blogs.sciencemag.org/pipeline/archives/2019/10/23/th... [2] https://search.sciencemag.org/?searchTerm=Lowe https://search.sciencemag.org/?searchTerm=Lowe alzheimers&order=tfidf&limit=textFields&pageSize=10&&
    - zosima 7y ago
      Different dementia rating scales may give different results because the study is underpowered for the effect, and/or Aducanumab is only effective in a subset of patients. There is something to suggest that the latter is the case. The MMSE score usually becomes low relatively late in dementia. The questions are easily solvable for most non-affected people [1] and the cognitive impairment must be quite severe for a significant drop. ADAS-COG on the other hand is more sensitive, especially in early disease [2]. It would be plausible then that patients in an early dementia stage treated with a higher dose, respond, while patients in a later disease stage can not be treated using Aducanumab. (Maybe it even has a paradoxical detrimental effect in late-stage dementia, who knows?) I would hope that they dig down in the data and find an objectively identifiable subpopulation where there appears to be significant effect, and rerun a phase III trial on that subgroup. Even if they only can treat a subset of Alzheimer patients, that's a very significant breakthrough. [1] https://www.bgs.org.uk/sites/default/files/content/attachment/2018-07-05/mini-mental_state_exam.pdf https://www.bgs.org.uk/sites/default/files/content/attachmen... [2] https://www.ncbi.nlm.nih.gov/pubmed/12079572 https://www.ncbi.nlm.nih.gov/pubmed/12079572
sampo 7y ago
> All in all, there is no question that amyloid-beta is a key part of Alzheimer. You are responding to an article that very much puts it into question. You should read the article, it's not long. But it'll give some quotes: “The amyloid hypothesis has been one of the most tragic stories [in] disease research,” said neurobiologist Rachael Neve of Massachusetts General Hospital. And Making it worse is that the empirical support for the amyloid hypothesis has always been shaky.
- zosima 7y ago
  Rachael Neve seems to believe, that APP (amyloid precursor protein) has a pathological role through other means than its degradation to amyloid beta, suggesting instead that it may have receptor-like properties [1]. Mostly, as I understand it, from arguments of structural similarity. It's interesting, but I know of very little solid evidence pointing that way. (And the article I could find was quite old) I think suggestions that amyloid beta is part of cellular defense, and that its "designed" to stick to foreign bodies/pathogens, preventing them from spreading in the brain, is more interesting.[2] Early-onset Alzheimer would then be caused by amyloid beta being released erratically, while many of the late-onset Alzheimer cases might be driven by amyloid beta in addition to some underlying inflammatory process. The amyloid hypothesis, strictly interpreted: that amyloid beta is highly neurotoxic through plaque formation and is the sole cause of neurodegeneration, is also known to be dubious and not explain many observations very well. But that doesn't really challenge that amyloid beta plays a significant part in the pathology. [3] [1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1862818/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1862818/ [2] https://www.frontiersin.org/articles/10.3389/fnagi.2018.00224/full https://www.frontiersin.org/articles/10.3389/fnagi.2018.0022... [3] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4888851/#emmm201606210-blk-0001 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4888851/#emmm20...
atombender 7y ago
I like the work that Cortexyme is doing [1] here: But in 2019, evidence pointed the finger elsewhere: at Porphyromonas gingivalis, a type of bacteria involved in gum disease. Research ... suggests that beta-amyloid is a symptom, not a cause of the condition. Instead, it may be the toxins released by P. gingivalis, called gingipains, that result in the brain damage that brings on the disease. Cortexyme is currently conducting a clinical trial to test an experimental drug that block gingipains. Their hypothesis coincidentally also explains why Alzheimer prevalence differs among ApoE genotypes. (I posted the story earlier today [2], but it didn't end up on the front page.) [1] https://www.newscientist.com/article/mg24432613-800-im-testing-an-experimental-drug-to-see-if-it-halts-alzheimers/ https://www.newscientist.com/article/mg24432613-800-im-testi... [2] https://news.ycombinator.com/item?id=21917192 https://news.ycombinator.com/item?id=21917192

by zosima 7y ago

amluto 7y ago

Nothing says p-hacking like a drug company canceling a trial and then, a few months later, re-analyzing the data and magically finding a benefit.

zosima 7y ago
FDA usually puts some quite strict limits on the amount of p-hacking and general shenanigans they allow in clinical trials. But there's no approval yet, and no data to look at, so it's too early to judge. Anyway, usually Alzheimer trials are a jumble of noise. The cognitive measurements are imperfect and patients are improving and deteriorating randomly, taking a very different disease course than you'd expect from textbooks. Finding a signal in the data, even for a working drug can be quite hard. I could just as well believe that Biogen ended the trial hastily because all competitors had failed and they didn't do a thorough analysis.
- memling 7y ago
  Derek Lowe's comments on the reboot were understandably a bit skeptical: > [W]hile the new larger EMERGE data do look like there’s a dose-responsive effect in the primary measurement (the CDR-SB scale) and three secondary dementia scores, the ENGAGE data actually show worse than placebo for CDR-SB at the higher dose (see slide 19 in Biogen’s presentation). They’re also worse than placebo in the MMSE scoring, but both the ADAS-Cog13 and the ADAS-ADL-MCI scales show apparent dose-responsive improvement. I have no idea why four different dementia rating systems should give such different results, and to a first approximation, no one else seems to have a good idea, either. So we’ve got the EMERGE/ENGAGE difference to explain (you’ve seen Biogen’s explanation, higher dosing later in the trial), and the variable results within the ENGAGE trial itself. And who are these high-dose patients in the ENGAGE study that Biogen says line up with the EMERGE trial? Ed Silverman at Stat quotes an analyst saying that it looks like fewer than 100 patients out of the whole set. That reminds me a bit of Biogen and Eisai’s tease of “We’re seeing positive data!” during the clinical trials, which also turned out to be based on a smaller number of patients than one would want when making a call on Alzheimer’s efficacy.[1] He's written a fair amount about Alzheimer's,[2] and it's been an informative and useful commentary (at least for me). [1] https://blogs.sciencemag.org/pipeline/archives/2019/10/23/the-return-of-aducanumab https://blogs.sciencemag.org/pipeline/archives/2019/10/23/th... [2] https://search.sciencemag.org/?searchTerm=Lowe https://search.sciencemag.org/?searchTerm=Lowe alzheimers&order=tfidf&limit=textFields&pageSize=10&&
  - zosima 7y ago
    Different dementia rating scales may give different results because the study is underpowered for the effect, and/or Aducanumab is only effective in a subset of patients. There is something to suggest that the latter is the case. The MMSE score usually becomes low relatively late in dementia. The questions are easily solvable for most non-affected people [1] and the cognitive impairment must be quite severe for a significant drop. ADAS-COG on the other hand is more sensitive, especially in early disease [2]. It would be plausible then that patients in an early dementia stage treated with a higher dose, respond, while patients in a later disease stage can not be treated using Aducanumab. (Maybe it even has a paradoxical detrimental effect in late-stage dementia, who knows?) I would hope that they dig down in the data and find an objectively identifiable subpopulation where there appears to be significant effect, and rerun a phase III trial on that subgroup. Even if they only can treat a subset of Alzheimer patients, that's a very significant breakthrough. [1] https://www.bgs.org.uk/sites/default/files/content/attachment/2018-07-05/mini-mental_state_exam.pdf https://www.bgs.org.uk/sites/default/files/content/attachmen... [2] https://www.ncbi.nlm.nih.gov/pubmed/12079572 https://www.ncbi.nlm.nih.gov/pubmed/12079572

sampo 7y ago

> All in all, there is no question that amyloid-beta is a key part of Alzheimer. You are responding to an article that very much puts it into question. You should read the article, it's not long. But it'll give some quotes: “The amyloid hypothesis has been one of the most tragic stories [in] disease research,” said neurobiologist Rachael Neve of Massachusetts General Hospital. And Making it worse is that the empirical support for the amyloid hypothesis has always been shaky.

zosima 7y ago
Rachael Neve seems to believe, that APP (amyloid precursor protein) has a pathological role through other means than its degradation to amyloid beta, suggesting instead that it may have receptor-like properties [1]. Mostly, as I understand it, from arguments of structural similarity. It's interesting, but I know of very little solid evidence pointing that way. (And the article I could find was quite old) I think suggestions that amyloid beta is part of cellular defense, and that its "designed" to stick to foreign bodies/pathogens, preventing them from spreading in the brain, is more interesting.[2] Early-onset Alzheimer would then be caused by amyloid beta being released erratically, while many of the late-onset Alzheimer cases might be driven by amyloid beta in addition to some underlying inflammatory process. The amyloid hypothesis, strictly interpreted: that amyloid beta is highly neurotoxic through plaque formation and is the sole cause of neurodegeneration, is also known to be dubious and not explain many observations very well. But that doesn't really challenge that amyloid beta plays a significant part in the pathology. [3] [1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1862818/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1862818/ [2] https://www.frontiersin.org/articles/10.3389/fnagi.2018.00224/full https://www.frontiersin.org/articles/10.3389/fnagi.2018.0022... [3] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4888851/#emmm201606210-blk-0001 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4888851/#emmm20...

atombender 7y ago

I like the work that Cortexyme is doing [1] here: But in 2019, evidence pointed the finger elsewhere: at Porphyromonas gingivalis, a type of bacteria involved in gum disease. Research ... suggests that beta-amyloid is a symptom, not a cause of the condition. Instead, it may be the toxins released by P. gingivalis, called gingipains, that result in the brain damage that brings on the disease. Cortexyme is currently conducting a clinical trial to test an experimental drug that block gingipains. Their hypothesis coincidentally also explains why Alzheimer prevalence differs among ApoE genotypes. (I posted the story earlier today [2], but it didn't end up on the front page.) [1] https://www.newscientist.com/article/mg24432613-800-im-testing-an-experimental-drug-to-see-if-it-halts-alzheimers/ https://www.newscientist.com/article/mg24432613-800-im-testi... [2] https://news.ycombinator.com/item?id=21917192 https://news.ycombinator.com/item?id=21917192