Skip to main content

Alzheimer's Disease

Another Prospective Alzheimer’s Trial

I wrote here about a prospective Alzheimer’s trial that’s starting soon among a population in Colombia, and now comes word of another large effort along the same lines. DIAN, the Dominantly Inherited Alzheimer’s Network, will test several current Alzheimer’s candidates in groups of people around the world with genetic mutations that make them susceptible to the disease. The hope is that these different mutations will provide a fast-forward-button look at the progress of Alzheimer’s in the general population, and help to settle the question of which mechanisms (if any) are appropriate to fight it. They’ll go two years of clinical observation (memory tests and brain imaging), and then the plan is to switch everyone to the most efficacious therapy and continue monitoring for real-world benefits.
Currently, it looks like there will be three candidates, with two from Eli Lilly: their beta-secretase inhibitor LY2886721, and their circulating-amyloid antibody Solanezumab, currently the subject of controversy about its efficacy or lack of same. The Roche antibody gantenerumab, which appears to bind more to amyloid that’s already precipitated, completes the trio.
This is an excellent idea, and I’m very glad to see so much work being done on prospective trials like these. There’s always the danger that working in genetic-mutation populations will give you an answer that’s not generally applicable, but I think that we know enough about the specific mutations to make a call on that, should anything stand out. The worry, naturally, is that nothing will stand out. The DIAN trial and the Roche crenezumab trial in Colombia are all aimed at various parts of the amyloid hypothesis, which has been the dominant strain of thought in Alzheimer’s etiology for decades. If nothing distinctive comes out of these efforts, that hypothesis will have taken some major hits – but they’ll have to be major hits to damage it in the first place.
The best result will be if something looks useful in preventative or early-stage Alzheimer’s. Second best would be a painful realization that the amyloid hypothesis is insufficient. And way down at the bottom would be a bunch of “Well, maybe. . .” clinical data showing that some of the agents seemed to help some of the patients, some of the time, to an extent, but maybe not enough to be effective by real-world standards. Anything but that, please.

10 comments on “Another Prospective Alzheimer’s Trial”

  1. Mohamed Shoreibah says:

    We need a reset in this area. Better to try either of two new oral compounds developed recently J147 or Dihexa (developed by two different groups). Both are BDNF mimetics or inducers and their pre-clinical data looks great. I would favor spending money on these oral (and substantially cheaper leads) with a novel mechanism of action than another set of amyloid-targeting flops.

  2. luysii says:

    Actually 3 prevention trials are in the works of which DIAN is one [ Science vol. 337 pp. 790 – 792 ’12 ].
    API (Alzheimer Prevention Initiative) will treat 300 members of Colombian families including 1000 carriers of a mutated PSEN1 (G280A). They’ll use Crenezumab, presumably an anti-Abeta antibody.
    DIAN (Dominantly Inhertied Alzheimer Network) — 240 members of families with early onset Alzheimer’s, of which 60 have a mutation on one of 3 genes (PSEN1, PSEN2, APP) — they will try 3 antiamyloid therapies (undetermined as of this writing 8/12). [ Science vol. 337 p. 1488 ’12 ] Initial analyses in the DIAN cohort ‘suggest’ that Abeta42 levels in the CSF begin to decline as early as 25 years before expected symptoms onset. Amyloid deposits begin (detected by the Pittsburgh compound) and increased tau in the CSF — begin 15 years before symptom onset. Cerebral hypometabolism and subtly impaired episodic verbal memory seem to begin 10 years before expected symtpoms in this cohort. It isn’t clear if this also happens in the 99% of people with sporadic Alzheimer’s.
    A4 Antiamyloid Treatment of Asymptomatic Alzheimer’s. 1500 healthy seniors, including 500 with amyloid positive brain scans. They will use an as yet unchosen Rx.
    Concentrations of beta amyloid in CSF dip ‘up to’ 25 years before the onset of Alzheimer symptoms.
    This is the rationale for trying out therapies BEFORE people get into trouble. Hopefully they won’t shoot their mouths off based just on biologic markers — remember CD4 counts and AZT? The best will be prevention of dementia, proof of which will take decades.

  3. gippgig says:

    Anyone looking at PERK inhibitors?

  4. bank says:

    I’ll go ahead and predict that these trials will fail. The antibody approach has already failed in several clinical trials, even in one using the same antibody (Solanezumab). The BACE inihibitor approach is aimed at reducing the production of Abeta, and thus amyloid, etc. However… we now know that Abeta levels _decline_ with disease onset, and so are not likely to be the agent responsible for worst effects of Alzheimer’s, at least after disease onset.

  5. steve says:

    Actually, you’re wrong about BACE. deCODE published a paper in Nature showing that people with mutations in APP that prevent BACE cleavage are strongly protected against Alzheimer’s and even against normal cognitive decrease with age. This clearly validates BACE cleavage as a target in Alzheimer’s.

  6. luysii says:

    Steve’s right. The paper is Nature vol. 388 pp. 96 – 99 ’12 (2 August), and accompanying editorial about it is pp. 38 – 39

  7. bank says:

    @steve and lyusii,
    I’m familiar with that Nature paper. However, the data measuring Abeta in the brain and CSF is far more extensive, and reproduced more often, than a single epidemiological study. Despite the arguments in favor of Abeta being the toxic agent in Alzheimer’s, it is clear that Abeta levels in the CSF are a) lower in people with Alzheimer’s and b) decline as the disease progresses.
    To quote [1]: “The most widely accepted explanation for the reduced CSF level of Ab42 in AD is that the aggregation of Ab into plaques (and thus retention in the brain parenchyma) results in less Ab being available to diffuse into the CSF”.
    However 1) plaques, while the focus of neuropathy are themselves not toxic, and 2) plaques do not get more extensive as the disease progresses. So the issue of amyloid toxicity remains unresolved, for the time-being. In fact, it is only clinical trials that are likely to resolve the debate, as mice are poor models for AD. This could be one useful result from the trials mentioned above, whatever the outcome.
    [1] Review:
    Fluid Biomarkers in Alzheimer Disease
    Kaj Blennow, Henrik Zetterberg, and Anne M. Fagan
    Cold Spring Harb Perspect Med doi:10.1101/cshperspect.a006221

  8. steve says:

    The data you site do not address the role of Abeta. Just because circulating levels in CNS fluid go down with progressing disease is not evidence that APP cleavage is not the central initiating event or that local amyloid accumulation in plaques is not the determinative factor. It just means that less leaks into the CNS fluid. The fact that people with a mutation that blocks the cleavage site are protected from Alzheimer’s is a lot more convincing, in my opinion.

  9. bank says:

    In the absence of good mouse models for the effect of APP in AD, the epidemiological and biomarker evidence for mechanisms in AD make the strongest case. Together they strongly implicate the region of APP that interacts with presenilin as key to AD. What happens afterwards is what is uncertain.
    If Abeta itself was the principal etiological agent, then one would expect that increased levels of Abeta would correlate with increased neurodegeneration. However, that is contrary to what is observed, at least with respect to Abeta in the CSF. If decreased CSF was correlated with increased plaque load, then one could argue that retention of Abeta in the brain is a better correlate for neurodegeneration than CSF Abeta. However, that is also not the case, as increased plaque load does not correlate very well with increased severity of disease. (Tauopathy on the other hand does.)
    Taken together, these data show that the question as to whether Abeta is the etiological agent is unresolved.
    The BACE inhibitor trial will be particularly interesting in this regard, since if the biomarker data is a true reflection of the underlying causes of AD, then inhibiting BACE will increase rather than decrease neurodegeneration, and possibly kill the notion of Abeta as being the toxic agent. Alternately, if the BACE inhibtor prevents AD, then the Abeta hypothesis will have been vindicated.
    I can only say that it’s a pity that we will have to wait several years for the outcome.

  10. Anonymous says:

    I like PERK also. However, one thing I noted is that people do exist with a Loss of Function mutation of PERK. They appear to be immune compromised to some degree. Not a great idea for an AD patient. Still, I wonder if anyone has done any epidemiological work on AD rates in these people….

Comments are closed.