. Rapid cognitive improvement in Alzheimer's disease following perispinal etanercept administration. J Neuroinflammation. 2008;5:2. PubMed.


Please login to recommend the paper.


  1. Overall, my comment is one of strong doubt because of the possibility that the observed improvement is a placebo effect, and because of the lack of double-blind information. That said, TNFα is a microglial-derived, secreted protein that is likely to be elevated in Alzheimer’s. Given that it is a known inducer of complement proteins such as C1q, it is a provocative idea that TNFα might be playing some role in enhancing neuroinflammation and perhaps even complement-mediated synapse loss in neurodegenerative diseases such as Alzheimer’s.

    Until a double-blind trial is run, we simply do not know whether TNFα antibodies will be useful for AD. Given that these antibodies are reasonably safe, I am all for doing a real trial to find out. One big caveat is that most antibodies cannot get across the blood-brain barrier.

    One more point: in clinical practice, elderly patients occasionally come to the emergency room with a bladder infection presenting as stupor or coma. The infectious agent or the inflammatory response to them depresses brain function, and these patients quickly become alert on antibiotics. Perhaps there is some inflammatory response in AD that, when suppressed, enables patients to suddenly do better.

    View all comments by Ben Barres
  2. Dr. Ed Tobinick has developed a revolutionary approach to deal with the neuroinflammatory damage which occurs in AD. He has administered Enbrel®, a soluble TNF blocker which is too large to cross the blood-brain barrier, perispinally to AD patients. The clinical results he reports are too dramatic to be overlooked by clinicians who plod along prescribing drugs that do not deal with the underlying pathology and which, at best, are only marginally effective in some individuals. Confirmation is urgently required to determine if, for the first time, there is a treatment for AD that really works. Dr. Tobinick reports almost immediate results so, at the start, a long-term trial with hundreds of patients is not needed. Let us hope that clinicians with access to patients will soon test the method and report whether or not they can confirm Tobinick’s results.

    Dr. Tobinick himself might want to provide control data by injecting placebo. As for our group, we did PET scans on rats injected by Dr. Tobinick with radiolabeled antibody. In this experiment, etanercept rapidly reached the CSF but not the brain parenchyma.

    View all comments by Pat McGeer
  3. I have been told that patients of colleagues received this treatment and it did not help them. I also have been told that the patients pay for it out of pocket, and that it is expensive. Last year the UCLA ADRC invited Dr. Tobinick to present his results to us. Dr. Tobinick’s presentation was for the most part a literature review on TNFα and inflammation. There was very little data and no formal study. Dr. Tobinick showed a videotape of a woman who performed poorly on simple directed tasks such as counting backwards from 10, then she was said to be treated, then she did better within 15 to 30 minutes. Since Amgen owns the drug, we called our contacts there, who said they had taken a look at this work and decided it did not merit follow-up. I asked Dr. Tobinick about Amgen—after all, if it worked, they would make money off it—and he said he presented to them and they didn’t follow up.

    There are several other issues:

    1. No controls for placebo or test/retest, nothing formal to show a real drug effect.

    2. A response so rapid that most TNF mechanisms could not be involved. It would have to be an immediate immuno-neutralization and signal transduction effect.

    3. Tony Wyss-Coray's panel finds low, not high, plasma TNFα as a diagnostic marker for AD (Ray et al., 2007). So a plasma effect would be in the opposite direction from normal but still might lower central TNFα. I think we can agree that most of the antibody would go to the periphery. This matters because TNFα can cross the BBB, so etanercept would lower peripheral TNF, which may be a significant source for brain TNFα and still have an effect. It is entirely possible that injection via another route would produce the same benefit in patients who have high peripheral TNFα.

    4. For the case report with an N = 1 and sudden improvement, it is hard to conclude that there is any real phenomenon to explain. But for the previously published open-label study with 15 patients, there are more interesting results to follow up on. In that study, they found evidence of a more gradual and continuing improvement in cognitive measures with no improvement of MMSE scores at 1 month, but then improvement at 2 months. That would be consistent with a more plausible, slower time course for an anti-TNFα effect.


    . Classification and prediction of clinical Alzheimer's diagnosis based on plasma signaling proteins. Nat Med. 2007 Nov;13(11):1359-62. PubMed.

    View all comments by Gregory Cole
  4. The role of TNFα in the brain is complex. Although there is a lot of indirect evidence that TNFα is detrimental in AD, this protein can be neuroprotective and has been shown to mediate synaptic scaling, a process that helps cells to readjust synaptic strength after prolonged activity (Stellwagen and Malenka, 2006). Given these diverse functions, we cannot yet say what increased or decreased levels in plasma or CSF mean for this disease.

    I think we should be cautious with inhibiting TNFα in AD, as AD is not an acute inflammatory disease.


    . Synaptic scaling mediated by glial TNF-alpha. Nature. 2006 Apr 20;440(7087):1054-9. PubMed.

  5. Dr. Ed Tobinick reports clinical improvement in AD patients administered etanercept, a macromolecular drug that interferes with TNFα activity. His recent Journal of Neuroinflammation paper reports the unexpected and intriguing finding that perispinal administration of etanercept leads to observable clinical improvement in AD patients within minutes. This case report of one patient follows up an earlier 2006 report in Medscape General Medicine. In that pilot study, 15 patients with mild to severe AD were treated once weekly with etanercept by perispinal administration for 6 months, and showed significant and sustained improvement in three different cognitive tests. Dr. Tobinick himself has raised several caveats related to the study, such as the possibility that the rapid effects are related to mood elevation or improved alertness, the lack of a control group treated with placebo, and the unconventional drug delivery method. The clinical observations are sufficiently intriguing to warrant analysis in a randomized, double-blind, placebo-controlled clinical trial.

    These initial clinical observations from a feasibility study are consistent with the hypothesis that cytokine levels and activity need to be tightly regulated in the brain, and that cytokine dysregulation might be a pathophysiology progression mechanism that can be targeted in new therapeutic development. The mechanisms behind how etanercept treatment influences cognition, especially the rapidity of the effects in the recent study, are not known and are one of many parameters that need to be studied in more detail. However, a number of testable hypotheses can be proposed and rapidly tested in preclinical animal models in parallel with planning for future clinical trials.
    The clinical observations also indicate the critical need for small molecule therapeutics that modulate cytokine production. This critical need is amplified when one realizes that current cytokine-modulating drugs are macromolecules, and using macromolecules as a therapeutic approach has a number of disadvantages for clinical use in chronic CNS disorders, such as high cost and inconvenient dosing regimens. Clearly, there is an unmet need for small molecule, orally active, brain-penetrant, anti-cytokine compounds to test as potential new classes of AD therapeutics.

  6. The media have been tripping over themselves about a single case report of what they have pegged as etanercept’s Lazarus-like effect in one patient. To add but one headline to the ones cited as examples in Gabrielle Strobel’s news story, the Methuselah Foundation quotes the author as saying, "the patients improve literally before your eyes" (Enbrel Works Fast Against Alzheimer's Disease?, accessed January 18, 2007).

    No matter how strong the underlying science or rationale of why an antagonist to TNFα might help, case reports tell so little and mislead so much. The circumstances of this one are particularly concerning, touting immediate and huge cognitive benefits using a perispinal injection process possibly covered by 17 patents owned by the author. The patient was treated just last October 30 for 7 weeks, with his final assessment on December 19 and journal publication the first week of January.

    Now 12 weeks later, would it be presumptuous to ask how the patient is doing? As reported, he was severely demented and confused before treatment, and after extrathecal infusion and being tilted head-below-his-heart for 5 minutes, improved markedly to a level of about moderate dementia in about 1 hour.

    I wonder what was the rush to publish this case when 1.5 years ago, the same authors published a 16-patient case series where they exampled one patient with a MMSE of 0 who had even greater effects than this most current case, i.e., a 34-point improvement on the Severe Impairment Battery, and apparently sustained over 6 months. What deadline were they trying to meet?

    They state in their case report that they had many similar experiences: “It should also be emphasized that rapid cognitive improvement following perispinal etanercept is not limited to the patient of the present report, but has, in fact, been commonly observed in multiple patients during the authors’ now more than 3-year clinical experience utilizing perispinal etanercept for treatment of probable Alzheimer’s disease.” Why didn’t they report these patients as well and earlier? How many other patients were treated and not reported? Did they have any failures, any non-responders? If you value one case, you have got to value them all.

    Uncontrolled case reports at best suggest hypotheses, regardless of how large the effects seem to be. If the efficacy of etanercept is even a fraction as great as that reported in this case or as the average 5-point ADAS-cog or 16-point SIB improvement over 6 months reported previously, then a simple and small double-blind placebo-controlled trial could be done very easily and quickly (preferably by independent investigators). The legal and ethical issues enumerated by Gabrielle Strobel should not be underestimated.

    Disclosure: Consultant with Wyeth, co-marketers with Amgen of Enbrel®.


    . TNF-alpha modulation for treatment of Alzheimer's disease: a 6-month pilot study. MedGenMed. 2006;8(2):25. PubMed.

    . Rapid cognitive improvement in Alzheimer's disease following perispinal etanercept administration. J Neuroinflammation. 2008;5:2. PubMed.

    . Perispinal etanercept for treatment of Alzheimer's disease. Curr Alzheimer Res. 2007 Dec;4(5):550-2. PubMed.

    View all comments by Lon S. Schneider
  7. While I am happy that the Alzforum linked to my blog, I don't feel that you sufficiently nailed the point down here ... that is, extraordinary "wow" headlines have no place in press releases about medical research. This is unfair to those with the disease, or to their caregivers. In this blogger's opinion, these authors lost all credibility when someone labeled their science with a headline like "Alzheimer's Symptoms Reversed In Minutes."—Dave Jensen, author, Sham vs. Wham

    View all comments by David Jensen
  8. Should elderly Alzheimer disease patients be exposed to off-label administration of a drug that has been shown to be largely unable to cross the blood-brain barrier (e.g., Tweedie et al., 2007), but which lowers the immune system, leaving these patients more vulnerable to infections? See etanercept prescribing information.


    . TNF-alpha inhibition as a treatment strategy for neurodegenerative disorders: new drug candidates and targets. Curr Alzheimer Res. 2007 Sep;4(4):378-85. PubMed.

    View all comments by Gerard Roberts
  9. Having spent some 25 years in industry focused on advancing therapies to treat patients with neurodegenerative diseases, I am very keenly aware of the critical importance for rigorous clinical trials.

    I do think, however, that Etanercept is deserving of further study. There is now a wealth of evidence from preclinical studies in acute and chronic neurodegenerative conditions that elevated TNF exacerbates injury. Dr. Clive Holmes demonstrated that the single best predictor of cognitive decline in AD patients is serum TNF levels (Holmes et al., 2009), a finding that has since been confirmed in larger numbers of subjects. These results suggest that targeting patients with elevated TNF levels might be a strategy for identifying what patients to treat with anti-TNF therapies. Such an approach, if combined with neuroimaging biomarkers for patient enrichment (ARF related news story) and molecular neuroimaging tools to assess neuroinflammation (peripheral benzodiazepine receptor radiotracers) suggests a path forward for designing innovative trials that address key hypotheses.

    Numerous comments highlight the concern about lack of BBB permeation of anti-TNF antibodies. This stands in contrast to the growing investment in industry now that is being placed on therapeutic antibody approaches to deliver disease-modifying therapies (reviewed by Paul, 2011).

    Note the reference on ClinicalTrials.gov to the study of the safety and tolerability of Etanercept in AD (STEADI-09), which is actively recruiting patients.


    . Systemic inflammation and disease progression in Alzheimer disease. Neurology. 2009 Sep 8;73(10):768-74. PubMed.

    . Therapeutic antibodies for brain disorders. Sci Transl Med. 2011 May 25;3(84):84ps20. PubMed.

    View all comments by Diane Stephenson
  10. Reply to comment by Diane Stephenson
    In general, the recent GWAS data provide a compelling rationale
    for some causal modulatory role for the immune system in the development of
    AD. But they don't clarify what this role is. It may be a positive role in
    amyloid clearance or a negative role related to neurodegeneration.

    The paper by Holmes et al. on which Diane Stephenson comments adds to a rationale to
    focus on TNFα, which is known to "orchestrate" inflammation in various
    models. That paper argues that TNFα plays a role in the rate
    of cognitive decline in AD, but most anti-inflammatory trials in AD patients
    have failed.

    The reports that some AD patients treated with anti-TNFα
    antibodies have had a favorable response are difficult to evaluate without a
    larger, more controlled clinical trial without any potential conflict of
    interest concerns.

    The issue with all anti-inflammatory interventions remains confounded by conflicting information about which population might
    benefit. For example, the ADAPT trial results now argue that persons with
    cognitive decline won't benefit from conventional NSAIDs like naproxen, while
    the paper by Holmes et al. suggests that anti-TNF therapy with antibody or even
    a small molecule like curcumin might actually help a subset of patients with
    established AD. This is really an empirical question.

    View all comments by Gregory Cole

Make a Comment

To make a comment you must login or register.