Skip to Content

Eli Lilly’s Alzheimer’s Antibody Does Not Work

After years of work and untold amounts of money, Eli Lilly’s Alzheimer’s antibody, solanezumab, does not work. It does not help Alzheimer’s patients. No matter how many times you run the Phase III trials, or in which patient populations, they do not get better. No matter how many hopeful press articles you might have seen on its prospects and no matter how many statements you might have seen from the company itself, no matter how you may try to spin the results, it does not do anything useful when you give it to actual Alzheimer’s patients. The program is over. It is done, because it does not work.

I’m so adamant because Lilly has been banging away on this drug for years now, hugely and expensively, which in one way does them some credit. Alzheimer’s is an insanely risky field to work in, what with the (basically zero) success rate, and the company deserves a lot of credit for taking it on. But what they do not deserve a lot of credit for is the way that they’ve been giving people to understand that yes, there’s hope, and yes, it’s actually working, and yes, the next trial could really be it.

And the press, at least a lot of the press, went along with this (especially in Britain). But over here in the US, we had “glimmer of hope“, “raised hopes“, “exciting possibilities“, and “will soon offer hope“. The most irresponsible headlines, though, were in the UK press, for reasons I’ve never quite understood. Meanwhile, during all this glimmering and hoping and excitement, most people who actually follow the field were in “You know, that doesn’t seem to have done much of anything” mode. I’m not criticizing Lilly for going ahead and seeing if solanezumab actually worked, although you could argue that this was an expensive long shot. But not enough people got the word about how long a shot it really was.

This news is also bad for Biogen, because they have a competing amyloid antibody in the works, and honestly, so far it’s not making anyone very hopeful, either. Antibodies are all different, so you can’t be sure, but this can’t be a good sign, and I (for one) will be very, very surprised if they get any convincingly good data when it’s all over. We’ll see.

Finally, this news is yet another whack at the entire beta-amyloid hypothesis for Alzheimer’s therapy. One of the key things that we might get from this latest Lilly trial is whether they saw signs of amyloid clearance/decreased amyloid aggregate formation in their patients. If they did, and if these patients didn’t get any better – and they didn’t – then where does that leave us? Trial after trial, mechanism after mechanism, and things always come up short. At some point, we’re all going to have to stop making excuses (not potent enough, bad PK, not selective enough, not the right patient population, didn’t run long enough!) Everyone’s always known that the entire beta-amyloid hypothesis could be wrong – a hypothesis always can be wrong. But facing up to what that means has been another thing entirely. You may not think that we’re at that point yet, and I can’t quite say that we are, but at some point we’re going to be at that point. Right?

99 comments on “Eli Lilly’s Alzheimer’s Antibody Does Not Work”

  1. Petros says:

    I was at a Lilly site for a conference at the time they announced this trial was going ahead despite the previous failure. The Lilly guys were all remarkably gung ho for it.

    Your last point is well made. But if no effects on amyloid were seen there may be one or two more attempts to batter at this brick wall!

  2. Barry says:

    Science moves forward by forming hypotheses and falsifying any of those that we can. What we can’t falsify experimentally we keep as provisional truth, on which we can build. To test he beta-amyloid hypothesis experimentally required inventing and delivering an agent (mAb in this case) that clears it from the living human brain. That in turn required an animal model, and a lot of protein engineering, and a clinical trial. If indeed they can show that the mAb cleared the plaques (will that require autopsy? Or can they be imaged in vivo?) most scientists will score the beta amyloid hypothesis falsified/wrong.
    But even then, we don’t always get a clean kill. Most of us deemed CETP disproven as a hypercholesteremia drug target with the failure of torcetrapib, but still, Merck jumped back in to fail at the same game.
    So are there standards on which we can agree for when a potential drug target is disproven?

    1. Arnab says:

      This is epic – ” What we can’t falsify experimentally we keep as provisional truth, on which we can build.” You should Print and Frame it .

  3. Gretchen says:

    It’s the oligomers that are toxic. The fibrils may be a sign you have lots of oligomers but not actually damaging. In fact, by sequestering oligomers, they might actually be protective, just as virus vaults sequester damaging viruses and keep them from harming.

  4. Anon says:

    The problem with this trial is that it ended in November. There must be a magic month somewhere. 11 more trials with this drug to go.

    1. Mark Thorson says:

      No, no, the problem was they started in these subjects too late. You need to start when amyloid accumulation begins — 10 years or more before symptoms appear. That’s why it didn’t work in these people. They were too far along the path to dementia.

      1. ScientistSailor says:

        Yes! The ‘Window’ hypothesis is that Ab increases years (~10) before phospho-Tau increases which occurs years before cognitive decline. Thus to really test the amyloid hypothesis, you would have to treat people 10 years before they have any symptoms. How do you run that trial you ask? You do it a Colombian family that is predisposed to early-onset Alzheimer’s:

        None of the trials that have been run so far even come close to testing the amyloid hypothesis

        1. steve says:

          Maybe they need to start with newborns and just follow them for 80 years…. Boy, the amyloid guys just never give up. What will the excuse be when the BACE inhibitors fail?

          1. ScientistSailor says:

            Steve, I don’t consider myself an ‘amyloid guy’ I just think they hypothesis should be tested properly.

          2. Hap says:

            Is there a way to do one of those, though?

            Sometimes I wonder if the amyloid hypothesis of Alzheimer’s is the string theory of medicine.

          3. Dolph says:

            Stirring a brain that is already puréed won’t help. It’s pretty much as simple as that. And just take it as a hint, this won’t change in the future. If there even is a practical way to “treat” Alzheimer’s it will be some kind of pretty early preventive step.

          4. scandriel says:

            Dolph: alzheimers is not a static condition. Recovery may not be an option but halting the progress would be easy to see and worth a lot.

        2. Mark Thorson says:

          Or you could test it in people with Down’s Syndrome, who frequently (but not always) develop early-onset AD. This is believed to be because the gene for the amyloid precursor protein is on chromosome 21, the chromosome triplicated in DS.

  5. MF says:

    (Disclaimer: non-scientist, asking a “blue sky” hypothetical, so yes I know this would be an ethical nightmare to try to implement, what with issues of prior consent and the like. My question is one of pure research utility, not feasibility)

    Would our understanding of Alzheimer’s Disease, and ways to address it, be accelerated if we could examine and experiment on the brains of living sufferers of the disease?

    The idea would be to recruit a large group of volunteers who have risk factors for Alzheimer’s who sign prior consents to be used as “living labs” for neurosurgical investigations and drug tests, should they develop the condition. The program would aim to keep them as comfortable as possible, obviously, and the compensation would be that their families and estates would be relieved of the burden of their care.

    With a predicted wave of dementia and neurological impairment cases in an increasingly elderly population, we certainly wouldn’t lack for experimental subjects (even if only 1/500 were willing). Would this help our understanding or not?

    1. Hap says:

      I think people only know whether someone has developed Alzheimer’s based on autopsy; it would be hard to know who had developed it from knowing only that they had dementia and thus to know what you were treating (it might give you information on dementia treatments, which might be useful).

      1. Mark Thorson says:

        You can image amyloid accumulation in the brain with florbetapir, and that’s what Lilly was using to screen people for the Expedition III clinical trial. This is not cheap — you need a cyclotron to make the radioactive fluorine, a synthetic chem lab to make the fluorine into florbetapir before it decays (half life is 110 minutes), and a PET imager to see where the amyloid is. Florbetapir only labels amyloid fibrils, not oligomers, so it’s not exactly a certain diagnosis of Alzheimer’s even among the believers in Amyloid Cascade Hypothesis 2.0.

        1. Anon says:

          What’s the point of imaging something that we now know doesn’t correlate at all with efficacy?

          So now, not only are the amyloid drugs useless, so are the imaging agents.

          1. Mark Thorson says:

            We don’t know that. We only know that solanezumab doesn’t work. We don’t know that florbetapir imaging is not predictive of developing AD. There’s good reason to believe it is predictive, though it’s also reasonable to be skeptical.

  6. Barry says:

    It seems that in the Alzheimers field at least, we have turned science on its head, acting as if a successful drug trial would “prove” the amyloid hypothesis, but refusing to discard the hypothesis despite all experimental falsification

    1. Mark Thorson says:

      No, they just didn’t do the trial right. Just give me another $100M and I’ll do it right. I’ll prove solanezumab works! Don’t give up now when success is in sight!

      1. steve says:

        On what plausible basis could anyone say that “success is in sight”? We’re getting into political pundit territory from the last election now.

        1. Hap says:

          I’m thinking he was being ironic.

          Success is in sight if you’re The Man With The X-Ray Eyes, perhaps.

          1. Mark Thorson says:

            Another hundred mill and we can do it right
            treat all day and test all night
            not another candidate left in sight
            in amyloidland . . .

        2. Dr CNS says:

          If we had only used the degu model….

  7. Lane Simonian says:

    The attempts to save the amyloid hypothesis are similar to Eli Lilly’s attempts to save solanezumab. They serve to kick the can down the road for another couple of years without any positive outcome.

    Amyloid oligomers may be more toxic than amyloid plaques but the c-terminal fragment of the amyloid precursor protein and excess amounts of the amyloid precursor protein itself may be more toxic than amyloid oligomers. And all cause oxidative stress (amyloid plaques in blood vessels at least). Ironically, lessening amyloid in and around neurons may increase amyloid plaques in and around blood vessels, resulting in a series of severe adverse effects including brain swelling and hemorrhaging–as has been seen in various trials using amyloid antibodies.

    The purpose of starting early is not to prevent various forms of amyloid from forming, the purpose of starting early is to limit oxidative stress. And there are a multiple of other factors that cause oxidative stress other than amyloid (in any form) so starting early at best may delay the onset of Alzheimer’s disease and slow down its early progression.

    Almost completely lost in the media were the results announced by Anavex yesterday. Professor George Perry who is a leading advocate for the nitro-oxidative stress hypothesis for Alzheimer’s disease assessed the results as follows:

    George Perry, PhD, Dean and Professor at the University of Texas at San Antonio and Editor-in Chief of the Journal of Alzheimer’s Disease, commented, “Although this is an open label study with 32 patients, I have never seen mild-to-moderate Alzheimer’s patients maintain near baseline cognitive and activities of daily living function and positive correlation with all other measures over a 41-week trial period in any prior study with an approved or experimental drug. It is quite plausible that complex CNS diseases like Alzheimer’s may require a comprehensive approach, including restoration of cellular homeostasis.”

    Anavex 2-73 more effectively inhibits excitotoxicity via NMDA receptors than either Aricept or Namenda. And this along with specific peroxynitrite scavengers is most likely the key to treating Alzheimer’s disease.

    1. Me says:

      “The multicenter Phase 2a clinical trial of ANAVEX 2-73 consists of two parts and a total of 32 mild-to-moderate Alzheimer’s patients. PART A is a simple randomized, open-label, two-period, cross-over between oral (30mg/50mg) and IV (3mg/5mg) administration, adaptive trial lasting up to 5 weeks for each patient. PART B is an open-label extension for an additional 52 weeks. Initially planned for 26 weeks, PART B was extended to 52 weeks as a result of requests from patients and caregivers”



  8. Kelvin says:

    Just 17 more Phase 3 trials with this drug and they should be able to achieve statistical significance at p < 0.05.

    1. Dr CNS says:

      Statistically speaking, one could divide 1 billion dollars worth of clinical work in 20 different studies and get one that is statistically significant.

      Why is the success rate in clinical trials in AD much lower than that random 1 in 20?

    2. Metacelsus says:

      Nope. You’ve fallen victim to the Gambler’s Fallacy. The expectation would be 20 more.

      1. Anon says:

        Tell that to Lilly.

  9. Russ says:

    Well, there is always the question of whether sola actually engages the target at all. This trial may not ever have been a useful test of the amyloid hypothesis.

    (Disclaimer: I’m one of the co-authors)

    1. Mark Thorson says:

      That’s very interesting, but you synthesized all of your antibodies? You don’t think there might be a difference between your solanezumab and Lilly’s solanezumab?

  10. anon says:

    Extremely small n, but the observation that you could have a brain that would be diagnosed as full-blown AD without actually have any cognitive impairment is intriguing.!/4071/presentation/19400 There are a lot of things very uncertain about our understanding of AD.

  11. loupgarous says:

    “The most irresponsible headlines, though, were in the UK press, for reasons I’ve never quite understood.”

    Let me help you here, Derek, having read British newspapers over breakfast for three months while I worked for my CRO’s UK branch… irresponsible headlines in the UK are because… British journalism,

    Your catch on Microsoft UK’s breathless annouincement that the oncologists could go home, the guys who gave us the Zune and Windows Vista would sort the cancer thing out for us in no time aside, British journalism has a few islands of stability (the Times and Daily Telegraph, for example) in a huge morass of tabloid-ish, slanted crap.

    My favorite example of this in what preens itself as the go-to place for sci-tech news in the UK and Australia is “Cameron denies personal Trident missile firing app, which amplified this prosaic written question for Mr. Cameron and his curt response:

    “Tom Blenkinsop: To ask the Prime Minister what discussions he has had with the Minister for the Cabinet Office about the development of a personal iPad application for his use”

    The Prime Minister: None.”(a direct quote from the Hansard for that day) into an entirely fictional discussion of non-existent nuclear command and control apps for Prime Ministerial tablet computers.

    There was some genuinely interesting discussion buried far below the lede for that bit of fluff masquerading as journalism; that any such app could conceivably be developed for 20,000 UKP – and even that’s mildly science-fictional to someone who knows what billable hours for coders run in the UK.

    1. Da Vinci says:

      You are overgeneralising. British TABLOID journalism (Mail, Express, Sun etc) is completely and utterly abysmal (and actually quite dangerous), but the BROADSHEET journalism (Guardian, Times, Telegraph etc) is pretty good. They all have the usual cluelessness about science that all papers have, but the headlines do differ.

    2. Andy Extance says:

      I’m curious about which (UK) headlines about Solanezumab people consider to be most irresponsible? I was intrigued that one of the ‘street of shame’ that Jacob Spieth highlights and Derek links to is actually written by a professor of neuroscience.

  12. anonymouse says:

    AD is a huge problem and kudos to Lilly for pushing this as hard as they have. I have zero to do with Lilly and I commend them for their commitment and bravery, we all should. The amyloid hypothesis is perfectly reasonable but it’s a hypothesis. Someone needed to do the experiment. It was an extremely costly bet and its bad for all of us that they didn’t succeed. The world needs success in AD. Punishing folks for trying is not going to get the hard problems solved.

    1. Hap says:

      I assume people are happy that they were willing to try. It’s just that, with limited resources, you want to try where you either are most likely to succeed or the most likely to get something substantial from failing. I don’t think it’s clear that solanezumab tested the amyloid hypothesis well, and it didn’t seem to work well in its other trials. No one knows everything, but this drug didn’t look like a good way to spend lots of money, because it doesn’t seem like people know anything more after it failed than before, and it was likely to fail. That’s probably what bothers people (or at least bothers me).

      The failure of solanezumab may also make it harder to get money to test the amyloid hypothesis more accurately, and for Lilly to try again with other potential Alzheimer’s drugs. Other than the people who got paid to run the trials, this was a really bad failure, and it was expected to be so, which makes it worse.

      1. Mark Thorson says:

        I don’t think anti-amyloid pharmaceuticals are anywhere near done. Amyloid Cascade Hypothesis 2.0 is still leading all of its competitors. Any competitor needs to explain why all of the mutations associated with early-onset AD are mutations which affect the amyloid precursor protein itself or the enzymes involved with amyloid processing.

        On the other hand, the current version of ACH 2.0 says that amyloid production is unaffected in AD, but amyloid clearance is impaired. It does appear to be true that there’s a lot of amyloid being made every day in the normal brain, the turnover is very high (about 2 weeks to replace it completely), and AD is not caused by increasing amyloid production (unlike our transgenic rodent models which have insanely high amyloid production).

        But if impaired clearance is at the root of the problem, why do we not see any mutations associated with amyloid transporters in EOAD? P-glycoprotein has been indicated as an amyloid transporter. There are drugs which strongly affect P-gp expression — why do we not see these drugs modulating AD? Explaining these contradictions will probably lead to ACH 3.0.

        1. Lane Simonian says:

          “A hypothesis that remains unproven yet catches the collective imagination can become, with the passage of time, so seductive that it dominates peer review opinion and arrests the development of alternative ideas. Such is the case for the amyloid hypothesis of AD. From the mid-1980s [1,2] this hypothesis began to give focus and excitement to what had been an unstructured research field with dozens of complex and unrelated theories [3], none of which dominated. It became a simple and effective way to describe AD pathogenesis to funding bodies, pharmacological companies, and the public at large.”

          Because genetic mutations lead to increased amyloid production in early onset Alzheimer’s disease does not mean that amyloid is the direct cause of Alzheimer’s disease. Gamma secretases actually made things worse, Lilly’s amyloid monomer antibody did not slow down the progression of the disease, Biogen’s amyloid oligomer and plaque antibody resulted in no slowing down in the loss of cognition over placebo at six months (although between six months and one year there appeared to be some slowing of progression). We are in essence down to BACE1 inhibitors as far as the amyloid hypothesis is concerned, and these are iffy as well since excess amounts of amyloid precursor proteins themselves can cause neuronal cell death.

          The key again is this:

          C-terminal fragment of amyloid precursor protein induces astrocytosis.

          Furthermore, conditioned media derived from CT105-treated astrocytes enhanced neurotoxicity and pretreatment with NO and peroxynitrite scavengers attenuated its toxicity.

          This explains why every effort to remove amyloid or stop its production has failed so far, but certain drugs and natural products that have partially reversed oxidative damage have produced promising results.

          1. Anonymous says:

            I understand your hypothesis, but there is one thing I don’t understand. It is known that the concentration of glutathione, the human body’s own antioxidant, is present in the brain of Alzheimer’s patients at concentrations exceeding 500 micromolar. Why would external antioxidants cure neurodegeneration (at concentrations less than 500 uM) whereas our own natural antioxidants are not effective ?
            There were over 400 clinical trials for AD between 2002 and 2012. Every time a small, moderately powered study showed an effect and was re-run as a well powered, double-blind study, the positive effects disappeared. I worry think will happen with your antioxidant studies as well.

          2. Anon says:

            Lane, you need to consider the direction of information in the genetic code:

            Genes (+ mutations) -> RNA -> proteins -> function (disease)

            Unless you can show that your proposed disease mechanism specifically causes mutations in amyloid-related genes, I think you have it the wrong way round.

        2. DN says:

          The genetic studies are pretty poor. They are only powered to detect alleles with a huge effect, and which create a disease that might not be true Alzheimer’s disease. A meaningful genetic study of Alzheimer’s would need at least thousands, and possibly millions, of DNA sequences, along with machine readable medical records.

          It would be like finding that Cushing’s disease ALWAYS includes elevated cortisol, and then trying to treat every hypercortisolemia disease with anti-cortisol antibodies, blockers of the last enzyme in its synthesis, receptor antagonists, disposal enzyme induces, etc. You would have to be a complete moron to try that.

          I consider it likely that amyloid-beta is the final effector hormone in a long cascade of signals. Abeta is probably what you get when the brain has decided to pull the pin out of its metaphorical grenade. The train of decisions before that is where we need to intervene.

          For that we need high-powered genetic studies, and in vivo comparisons of human brain hormones from birth to death, with special emphasis on progeria, encephalitis, brain cancer, early onset Alzheimer’s, and cognitively intact extreme elderly. None of which is being done today. Almost the entire research budget is being wasted.

          1. Lane Simonian says:

            Anon, it is more complicated than this. As I understand it, many amyloid precursor protein mutations leading to early onset Alzheimer’s disease make it easier for BACE1 to cut the amyloid precursor protein to form the c-terminal fragment (they don’t appear to increase BACE1 expression itself). Caspase-3 not only stabilizes BACE1 but also may participate in the processing of the amyloid precursor protein itself, so if you inhibit caspase 3 (or scavenge peroxynitrite that leads to caspase-3 activation) then you limit the effect of the mutation.

            Furthermore, the c-terminal fragment through g protein signalling is neurotoxic precisely because it increases peroxynitrite formation and caspase-3 activation. So if you scavenge peroxynitrite and inhibit caspase-3, you first limit the amount of the c-terminal fragment produced in the first place and then you limit its neurotoxicity.

            Curiously, many amyloid precursor mutations in early onset Alzheimer’s disease limit gamma secretase activitiy, and apparently as an upshot most amyloid ends up in blood vessels rather than in neurons. Cerebral amyloid angiopathy also leads to an increase in peroxynitrite and caspase-3 activity. So once again the strategy of scavenging peroxynitrite and inhibiting caspase-3 activity works.

            One last example, even though individuals with Down syndrome have an extra chromosome containing the gene that results in the production of extra amyloid precursor proteins not all of these individuals develop Alzheimer’s disease. Individuals with Down syndrome have high levels of hydrogen sulfide which is a peroxynitrite scavenger and in combination with other peroxynitrite scavenger this may delay or prevent the development of Alzheimer’s disease in some individuals with Down syndrome (although hydrogen sulfide itself is toxic to the brain).

            Several forms of amyloid cause oxidative stress at least at the early stages of Alzheimer’s disease, but they are one of several causes of that stress (perhaps more so in early onset Alzheimer’s disease than in late onset Alzheimer’s disease). But outside of oxidative stress, it is not clear that amyloid is neurotoxic at all. And if this is the case, then reduce and reverse some of the oxidative stress and you effectively treat Alzheimer’s disease.

  13. loupgarous says:

    There’s a thrifty Hoosier reluctance at Lilly (headquartered in Indianapolis) to throw out functional models that don’t pan out in Phase III.

    It was there with insulin lispro (a recombinant form of human insulin where lysine and proline are swapped on the C-terminal end of the B-chain – making the drug more bioavailable for post-prandial injection). There was a hypothesis that somehow this would reduce insulin resistance, and Lilly kept looking for that in Phase III trials – but it didn’t pan out. As with their late experience with solanezumab, it was an admirable desire to solve a vexing problem – and if it had panned out, Lilly and its customers would both have “won.”

    In “Humalog,”(Lilly’s trade name for insulin lispro) they did, however, have a nice, fast onset-of-action variant of human insulin, and even sold it mixed with “Humulin,” their own recombinant version of non-re-engineered human insulin to give patients and prescribers a lente version.

  14. Crocodile Chuck says:

    In the ’90’s a researcher & a physician propounded the view that gastric ulcers were caused by a bacterial infection.

    In 2005, they won the Nobel Prize for Medicine for their discovery.

    What if ALZ was caused by bacterial/fungi infections of the brain?

    1. Mark Thorson says:

      Baloney. Lots of people with AD get treated with antibiotics and antifungals for conditions unrelated to their AD. If an infection caused AD, there would be case reports of people whose AD had resolved after treatment. No such case reports exist. AD is not caused by an infection.

      1. loupgarous says:

        AD shares some pathophysiology with transmissible spongiform encephalopathies.

        It would explain the apparent failure (now that Merck’s beta secretase inhibitor has also failed in Phase III) of the beta-amyloid hypothesis if the amyloids characteristic of AD were a consequence of the disease process, not its cause.

        Some of the TSEs are also characterized by amyloid formation, but their cause is now recognized to be infection by misfolded proteins which cause the host’s cellular prion protein to be misfolded in the same way (just as two allotropes of the same crystalline substance can serve as “seed crystals” and propagate their respective structures in a supersaturated solution of the substance in a solvent, causing the crystalline substance to precipitate out of the solvent in two different ways – in World War II, this supposedly happened on the huge scale in a plant producing EDTA).

  15. Crocodile Chuck says:

    Please familiarise yourself with some physiology, in particular the blood:brain barrier, which prevents large molecules, including most antibiotics, from entering the brain.

    BTW, your attitude is precisely that of the gastroenterologists who derided Warren & Marshall’s theory in the 1990’s.

    1. Lane Simonian says:

      Brain infections (viral, bacterial, fungal, etc.) may be one of several causes of Alzheimer’s disease because these infections trigger peroxynitrite formation. And certain antibiotics may have a beneficial effect on Alzheimer’s patients.

      For example, minocycline which is a peroxyntrite scavenger:

      Curr Alzheimer Res. 2016 Aug 19. [Epub ahead of print]
      The anti-inflammatory role of minocycline in Alzheimer´s Disease.
      Budni J1, Garcez ML, de Medeiros J, Cassaro E, Santos-Bellettini T, Mina F, Quevedo J.

      1. Crocodile Chuck says:

        Thanks for the link, Lane.

    2. Mark Thorson says:

      Cefotaxime, ceftriaxone, chloramphenicol, clioquinol, doxycycline, minocycline, roxithromycin, and sulfamethoxazole are all antibiotics which cross the blood-brain barrier. Some of these are very widely used. Tens of millions of people have Alzheimer’s Disease. Where are the case reports of recovery from AD after treatment? They don’t exist. The notion is total baloney.

      1. Crocodile Chuck says:


        I’m afraid you’re missing the distinction between proximate and root causes

        If the latter is the ‘leakiness’ of the blood:brain barrier itself, then a 5 day course of antibiotics for something else [which incidentally kills bacteria in the brain within the interval] isn’t going to address the underlying problem

        ‘Leakiness’: Root Cause


        1. Lane Simonian says:

          You are welcome, Crocodile Chuck. The breakdown of the blood-brain barrier is in part the indirect result of peroxynitrite.

          “Peroxynitrite mediates nitric oxide-induced blood-brain barrier damage.”

          And this is part of the reason why certain antibiotics, such as minocycline, could potentially help in the treatment of several neurodegenerative diseases, at least in terms of slowing down the progression of these diseases.

          “Neuroprotection by Minocycline Caused by Direct and Specific Scavenging of Peroxynitrite*”

          Any treatment for the disease must indeed be long-term and continuous otherwise the processes behind the disease kick back in.

        2. Mark Thorson says:

          The recommended course of treatment for Lyme disease is 21 to 30 days of doxycycline. Hundreds of thousands of people in the U.S. have been treated for Lyme disease. Where are the case reports of resolution of AD after treatment for Lyme disease? You’d think there would be at least one.

          The infection hypothesis is not just wrong, it’s stupidly wrong. It’s peroxynitrite wrong. It ignores obvious refutation.

          1. Lane Simonian says:

            A clinical trial indicates that you are correct regarding doxycycline.


            But like some other antioxidants studies for Alzheimer’s disease, the trial probably did not use the best antibiotics.

            “Minocycline was also compared with doxycycline and rifampicin. These drugs were better PON scavengers than tetracycline and were able to prevent the oxidation of the DHR-123 indicator compound by PON. They might offer protection in some in vivo models if dosed high enough, but our data indicated that they were still ∼100-fold less potent than minocycline (not shown)” [from Neuroprotection by Minocycline Caused by Direct and Specific Scavenging of Peroxynitrite*]

      2. bacillus says:

        Hi Mark:
        No one made the connection between H. pylori and ulcers despite the fact that hundreds of millions of people with the disease took antibiotics at some point. For Hp it was the use of combination antibiotic treatment for at least a couple of weeks that led to cure. Moreover, some bacteria can go into a metabolically quiescent survival mode when exposed to antibiotics, only to bounce back once the therapy is withdrawn. Then there are bacteria such as M. tuberculosis that have a doubling time measured in days and can take 20+ years before causing overt pathology. For intracellular bacterial pathogens, antibiotics capable of accumulating within the infected host cell are required, but most commonly prescribed antibiotics lack this ability. I’m not saying that I believe the germ hypothesis regarding AD. I’ve been an infectious disease researcher for nearly 35 years, and have lived through many claims that bacteria are responsible for everything from arthritis to atherosclerosis. However, I do think you are writing this possibility (however minuscule) off with the clinical evidence at hand regarding a complete lack of correlation between antibiotic use and AD.

        As an interesting aside, Warren and Marshall had been trying for ages to culture bacteria from the stomachs of ulcer patients without luck. They only succeeded because they put some plates in the incubator during a long week-end holiday, and the extra day in the incubator was enough for visible colonies to grow.

  16. Lane Simonian says:

    Anonymous, the level of glutathione in Alzheimer’s disease is greatly reduced and this is a reflection of oxidation.

    As glutathione is depleted, external antioxidants must take its place. The most effective ones are methoxyphenols including ferulic acid, syringic acid, eugenol, and curcumin because they are good hydrogen donors. Curcumin is not well absorbed in the bloodstream and ferulic acid is not well-absorbed into the brain; neither seem to have much impact on cognition but ferulic acid has a positive effect on behavior in people with Alzheimer’s disease, dementia with Lewy bodies, and frontotemporal lobe dementia perhaps by lowering levels of norepinephrine (the evidence for curcumin in this regard is more limited). Use of multiple peroxynitrite scavenger appears to have a positive effect on cognition. One example is cbd oil (cannabidiol, THC, and terpenes including eugenol) although the evidence for this so far comes mainly from mice studies and case studies. Another example is panax ginseng (ferulic acid, syringic acid, p-coumaric acid, vanillic acid, and maltol). In a two year open label study with Korean red ginseng, improvements in cognition for participants with Alzheimer’s disease were seen at 24 weeks and were sustained for two years. Once the amyloid strategy fails; perhaps the antioxidant strategy will open up.

  17. loupgarous says:

    There’s a disquieting parallel in this enlightening discussion – what other poorly-understood condition manifests with amyloid fibrils in the brain (albeit in different areas, and often not with dementia as an early symptom)?

    The spongiform encephalopathies. And like that whole group of illnesses (scrapie, kuru, CJD, fatal familiar insomnia, GSS, BSE – and what what we think is probably BSE in human brains, “new variant CJD”), controversy between virologists and other researchers with decades of experience studying these diseases exists on the ultimate cause of the disorder. Stanley Prusiner got the Nobel in Medicine and Physiology for the “prion” theory, and sure enough, was able to passage prions from animal to animal to animal with inoculations of infectious material.

    But controversy remained for quite a while – Fields’ textbook on virology at one time had a chapter with two parallel hypotheses for the causes of the spongiform encephalopathies, one by Prusiner expounding the prion theory, one by Carleton Gajdusek presenting what he viewed as possible alternative theories held by the discoverer of scrapie-associated fibrils, and still other alternatives to the prion model.

    It’s hard not to see research in Alzheimer’s Disease’s ultimate and even proximate causes as being at roughly this same stage – not that any evidence exists at present that prions are involved, but that there’s this same lack of consensus and clear evidence about what’s happening.

  18. hn says:

    Can Lilly survive this?

    1. loupgarous says:

      Eli Lilly is huge, physically and financially. It hurts to have a drug development campaign that expensive not pan out at all, but this is an outfit with not one, but two Crays in its warren of labs for molecular modelling and such. And having their state’s governor be Vice-President-elect won’t hurt them one bit.

    2. Dr CNS says:

      Survive? Possibly.
      Layoffs? Surely.

  19. luysii says:

    Very sad. I began the year by going to a memorial service for a college classmate, fellow doc and friend who died of Alzheimer’s, despite 50 clinical papers evaluating drugs like captopril to his credit.

    So new hypotheses about Alzheimer’s causation, however weird, shouldn’t be dismissed. Here are two non chemical ones —

    Here’s another (about gemfibrozil) which is actually being studied at this point —

  20. Lewis Rumpler says:

    I think the failure of the Amyloid hypothesis is that it has never addressed the relationship of plaque removal to restoration or improvement in cognition. This is the primary clinical endpoint all alzheimer’s drugs are measured by.
    If you can’t induce neuronal stem cell proliferation/differentiation, neurogenesis, synatogenesis and reduced oxidative stress then I think the odds of seeing some semblance of measurable cognitive improvement is unlikely. This leads you to the domain of growth factors, a biology which is settled science but escapes investment due to lack of BBB permeant agonists and failed gene therapy approaches and there is little appetite (for good reason) of direct transcranial injection of said growth factors.

    1. Mark Thorson says:

      The notion that plaques cause the symptoms of AD is Amyloid Cascade Hypothesis 1.0. That is so 1990’s. We’ve moved beyond that. ACH 2.0 blames Abeta oligomers for causing the destruction responsible for the symptoms of AD. Plaques may in fact be protective against these oligomers, because they trap and sequester them outside of the neurons. Also, there are many people who die in old age without ever having exhibited signs of dementia who have brains full of plaques. Those plaques may have been what saved them.

        1. Mark Thorson says:

          It’s a long way from demonstrating Abeta has an antimicrobial effect to claiming that bacteria cause AD, and those researchers do not claim that. Abeta is toxic to nerve cells, so why wouldn’t it be toxic to other cells? It may be harmful to bacteria, but that doesn’t mean its function is to kill bacteria.

  21. Anon says:

    Sola failed, I think, not because the amyloid hypothesis is wrong, but for more traditional drug-failure reasons – the PK was poor, and the target (through the magic and infallible retrospectoscope) not optimal. I suspect that the “directionality of change” noted in the PR is actually correct and indicative; they just simply didn’t trap enough of the right species. As others have said, kudos to Lilly for at least trying.

    I think the Biogen antibody will turn out to be a better shot, since it has superior CNS access and is directed to the oligomer.

    1. Lane Simonian says:

      We may know the answer to the effectiveness of Biogen’s drug soon. Biogen is slated to report on 24 month data for aducanumab at CTAD. It effectively targets both amyloid plaques and oligomers, although apparently not amyloid plaques in blood vessels.

      I am not expecting good results. At six months there was very little difference between aducanumab at any dose and placebo in regards to cognitive function. At somes doses there was a significant difference at one year.

      This is highly unusual for an Alzheimer’s drug to only start showing effectiveness after six months.

      Several reporters have claimed that Biogen carried last observed results forward. This would have skewed the results in favor of the drug. Biogen denies that this was part of their methodology, so it is difficult to determine who is telling the truth. The 24 month results should add clarity as to whether aducanumab actually slows down the rate of cognitive decline.

      1. Anon says:

        “This is highly unusual for an Alzheimer’s drug to only start showing effectiveness after six months.”

        Not really. Why do the FDA want (and sponsors want to sponsor) two years on-drug? It is because arresting the decline in CDR-sum of boxes (or whatever) isn’t powered over six months unless you were to massively increase the sample size beyond what is financially and practically feasible. Running two year studies is a compromise; the ideal would be longer.

        It would be great if there would be some diagnostic (not that there ever may be) that predicts who is going to get AD *before they have any symptoms at all*. Since this isn’t the case in general (though apoE4-4/biomarker positive subjects might be), you look for very early changes in cognition and biochemistry – and still you are hoping that the observed damage that allows you to enroll isn’t already too late, whilst still being able to aim at a general label.

        If this were easy, it would have already been solved. That doesn’t mean there is no solution.

  22. Anon says:

    Given the near-perfect correlation between genetic mutations of amyloid and hereditary forms of AD, the amyloid hypothesis is undeniable. But so are the countless failures of drugs designed to target amyloid. So we must focus on possible reasons for this discrepancy:

    1. Pharmacokinetics. The toxic form of amyloid is completely inaccessible to all drugs tested thus far because it is buried deep within brain tissue, and perhaps even buried within cell membranes (e.g., forming non-specific pores) which antibodies and other drugs cannot penetrate. Do we really know that the drugs can disrupt the toxic form of amyloid within the critical areas of the brain?

    2. Timing. Amyloid is the initial trigger, or at least a significant risk factor for AD, but once triggered, progression of the disease is irreversible and/or independent of amyloid. Thus, amyloid acts like a seed, but then the disease process continues with or without amyloid, and no drug targeting amyloid will be able to stop the process. Once the milk starts to curdle, that’s it.

    Perhaps we need to focus on answering these two questions with well-designed scientific experiments, rather than just throwing more crap at the wall to see what sticks?

    1. Lane Simonian says:

      Part of this may be true. Amyloid may be one of the triggers for Alzheimer’s disease, but there are potentially many other ones (high glucose levels, high blood pressure, various pesticides, air pollutants, industrial solvents, mercury, chronic heavy smoking, psychological stress, chronic bacterial, viral, and fungal infections in the brain, etc.). The question is how much of the oxidative stress results from amyloid and how much of the oxidative stress results from something else. By removing various forms of amyloid do you prevent Alzheimer’s disease or just delay its onset and if the latter by how many years if any?

      “Neurotoxic Mechanisms Caused by the Alzheimer’s Disease-linked Swedish Amyloid Precursor Protein Mutation

      “Early Onset Alzheimer’s Disease and Oxidative Stress”

      ” Mutations in APP, PS1, and PS2 genes are causes for early onset AD. Several animal models have demonstrated that alterations in these proteins are able to induce oxidative damage, which in turn favors the development of AD. This paper provides a review of many, although not all, of the mutations present in patients with familial Alzheimer’s disease and the association between some of these mutations with both oxidative damage and the development of the pathology.”

      The second supposition is likely also true (although I am twisting it a bit): Alzheimer’s is not an irreversible disease and after a certain point it progresses independently of amyloid. The receptors which different forms of amyloid putatively work through (g protein coupled receptors and receptor tyrosine kinases) are damaged in Alzheimer’s disease. The progression of the disease depends on the following loop: peroxynitrite–NMDA receptors activation–peroxynitrite. To partially reverse the disease, it is likely necessary to disrupt the loop.

      “Mechanism of Oxidative Stress and Synapse Dysfunction in the Pathogenesis of Alzheimer’s Disease: Understanding the Therapeutics Strategies”

    2. Dr CNS says:

      Correlation is not causation. Even your “timing” comment exemplifies why that is the case.

      As Russ wrote above, IF LLY are still uncertain whether the drug engaged the target in the expected tissue, what can be concluded from this work – other than “didn’t work”?
      And if we cannot conclude anything from an experiment, was it worth doing?

      1. Anon says:

        True, normally correlation is not causation, but unless you are willing to ignore the flow of information in the genetic code, from genes to disease, then all hereditary information IS causation.

        1. Dr CNS says:

          Not sure what you mean by “normally” in a scientific context.
          I am not wiling to ignore the flow of information in the genetic code, more than I am willing to say that obviously that *ALONE* is not a tractable drug discovery strategy.

          Personally, I believe we need to start in the other end of the cascade: proteins and how their function is determined. Isn’t that where the flow of information take us, “normally”?

  23. Anon says:

    “Amyloid may be one of the triggers for Alzheimer’s disease, but there are potentially many other ones (high glucose levels, high blood pressure, various pesticides, air pollutants, industrial solvents, mercury, chronic heavy smoking, psychological stress, chronic bacterial, viral, and fungal infections in the brain, etc.). ”

    If any of these obviously accounted for more than a very small percentage of risk, then we would almost certainly be onto something. But they don’t, and so we aren’t. What seems unarguable is genetic risk, and more arguably cholesterol (both being though in the end epidemiological observations).

    1. Mark Thorson says:

      This study finds that high total cholesterol in later life (>70 years) is associated with reduced risk of dementia, but mentions other studies find increased risk when cholesterol is measured at mid-life:

      It mentions some possibilities for this apparent contradiction. Note that the brain and spinal cord are the richest organs for cholesterol in the body. The industrial source of cholesterol for cosmetics and vitamin D production is the brains and spinal cords of cattle.

    2. Mark Thorson says:

      This study suggests that some statins increase the risk of cognitive impairment, specifically the lipophilic statins. That would be consistent with the idea that depleting cholesterol is bad for the brain.

      1. Barry says:

        statins do more than drop circulating levels of cholesterol. But blocking hMGCoA reductase, they block prenylation of proteins. Statins that spend less of their time outside the liver show less tox, but the problem is inherent in the drug mechanism

  24. Lane Simonian says:

    Cholesterol is a controversial risk factor for Alzheimer’s disease, because cholesterol does not normally enter the brain from other parts of the body, but in cases where the blood-brain barrier is damaged it may be a risk factor for Alzheimer’s disease. There is varying degrees of support for the risk factors for Alzheimer’s disease that I listed; some have better evidence than others.

    My only point is that genetics is not the single determinant for Alzheimer’s disease, even in the case of early onset Alzheimer’s disease. It does not explain, for instance, why there is a variation in the onset of the disease even in early onset cases. It also does not explain why some people with Down syndrome, who have extra amyloid precursor proteins in their brain, do not develop Alzheimer’s at all.

  25. anonon says:

    If only blog postings or pithy tweets could halt/reverse the progression of AD. We would surely have abundant treatments.

    1. Dr CNS says:

      I agree – these comments surely suggest the existence of a fair number of ideas that are not always represented by the preponderant views.
      Unfortunately, most of the ecosystem running our industry acts based on consensus, and getting support for novel ideas is asphyxiating it to death.

  26. Paramus says:

    I have to agree with Derek, it is not the testing of the hypothesis that is the problem, but the hype put out by companies and the media. I feel very sorry for patients who raised their hopes base on irresponsible press releases and media coverage. It still continues, look at STAT ( Stock analysts are saying why Biogen’s Ab is different and why it may work! When did the world change so we now listen to stock analysts with vested interests in stead of the vast majority of the Alz scientific community that now is convinced the hypothesis is dead!

    1. Dr CNS says:

      Good question – but this is hardly breaking news.
      When was the last time an analyst was down on a certain stock? They play their game.
      If the parameter that defines success is your material gains in the short term, long term activities such as drug discovery are not within your scope.

    2. David Edwards says:

      Remember Nick Leesom? So convinced that his spread betting centred on the price of the Yen on the currency markets would yield pay dirt, that he ended up betting the entire bank on it? Only for the bank to go the same way as the Titanic, when it hit the iceberg of reality, in the form of the Yen nosediving after the Kobe earthquake?

      History is replete with examples such as this: individuals or organisations pursuing a cherished idea to the point of their own destruction. Though the stock market is over-represented here, precisely because it’s populated with people whose sole motivation is the making of a fast buck, no matter how malign or inimical to the rest of society that motivation ends up becoming, especially when pursued with the religious zeal more usually considered an aspect of the aetiology of fundamentalism.

      At least with the likes of Lilly’s pursuit of solanezumab, there was something resembling strategic thinking behind it. But my view is that the failure thereof has opened up more questions than answers. Such as “if this finding points to beta-amyloid being more akin to battlefield debris than the actual artillery exchange, then what fired the first shot?”

      This could become a tail biter in numerous ways. Just to hypothesise for a moment (and I’ll admit here to having nowhere near the expertise of the actual researchers in the field, but they might appreciate me voicing this anyway), what if, after decades of research, it transpires that the first shot was in utero? Such a finding, if it ever becomes empirically established, is going to be a nightmare. Not least from the ethical standpoint of how one informs a seemingly healthy young adult, that thanks to the quirks of his or her instance of developmental embryology, there’s a nasty medical time bomb ticking away inside that person’s brain.

      It’s at times like this, that the daunting responsibilities resting on the shoulders of researchers start to make themselves manifest. Careful navigation thereof not being assisted by the spreadsheet mentality.

  27. John Hood says:

    Antibodies don’t cross the blood brain barrier. It is unreasonable to expect a benefit to the Brian for a therapy that doesn’t reach the brain.

    1. Janex says:

      Depends on how they are administered. I’ve seen a lot of direct to brain antibody administration protocols lately.

      1. Mark Thorson says:

        Some drugs can get into the brain past the BBB via the intranasal route. Perhaps that should be tried with solanezumab and the other antibodies. Add 10% cocaine to ensure patient compliance. 🙂

        1. Mark Thorson says:

          It’s already being done in rodents, and it works.

  28. John Hood says:

    In drug discovery we target 20% bioavailability for an oral drug to deliver benefit. Roughly 20% of an orally administered drug must reach systemic circulation to reasonably expect systemic benefit. Roughly 0-0.5% of these antibodies reach the CNS yet we expect a CNS benefit – why?

  29. Janex says:

    My personal thought is that Alzheimer’s will wind up being like cancer with multiple different causes each requiring different treatments based. Until we can subdivide Alzheimer’s patients the way we can subdivide cancer patients, treatment strategies will continue to struggle.

  30. Barry says:

    “This is certainly not the end of the road for the amyloid hypothesis, and trials of other monoclonal antibodies targeting amyloid β—such as aducanumab, which showed more promise than solanezumab in early phase studies—are ongoing. But it does highlight the need for research in other mechanisms (such as tau, neuroinflammation, and treatments that target multiple pathways), and to harness the accumulated knowledge that has been collected from “failed” clinical trials.”

  31. Deborah says:

    My husband was in the Expedition 3 trial and completed the 18 months and was in the beginning of the 2 years “off label”. They just pulled everything on us and I noticed a huge drop off in his cognitive abilities. He had early onset and was considered mild when we started the trial. He did extremely well prior to it being pulled. We went to the neurologist last week and his scores are way down. They dropped more than they had dropped in the entire 2 years. Now none of the testing is being released to us by Lilly (not even if he was on the placebo). Obviously this information could help in deciding any future trials. This worked for him and after seeing this change I am completely frustrated. This drug is being made and there should be a compassionate allowance for people who have shown positive results and we should not be cut off from everything. His results could help others as well as us. In the past 2 months our quality of life has changed more than in the past 2 years. There was no consideration for participants who benefited. Just an abrupt halt.

Leave a Reply

Your email address will not be published. Required fields are marked *

Solve the math problem. * Time limit is exhausted. Please reload CAPTCHA.