Articles

Don’t Believe The Hype: Cholinesterase Inhibitors As A Treatment For Dementia

Don’t Believe The Hype: Cholinesterase Inhibitors As A Treatment For Dementia
I distinctly remember the day I attended a “drug lunch” (as a PM&R resident in New York City) to learn about the value of donepezil (aricept) for the treatment of dementia. I was astonished by the drug’s lack of efficacy – the graph displayed in the PowerPoint show demonstrated a 2-point improvement on the Mini Mental State Exam (MMSE), an effect that began after 6 months of donepezil use, and persisted for only 6 months after that. A 2-point difference on the MMSE has no clinical relevance of which I’m aware. The drug’s common side effects include: nausea, vomiting, diarrhea, loss of appetite, tiredness, drowsiness, trouble sleeping, or muscle cramps. That day I vowed never to prescribe the drug to my elderly patients.
Nonetheless, I was dumbfounded by the number of patients who came to the hospital already on the medication. Over and over again I heard the same story: “mom is becoming forgetful so our doctor started her on this medication to help her memory.”
When I asked if the family if they thought the medicine helped, the response was equally predictable: a shrug and then “what else can we do?”
Here we have a classic example of a medical problem with no satisfactory treatment or cure – and a desperate desire on the part of patients and family members to do something – anything – about it. Many times people in these predicaments turn to alternative medicines, herbal supplements and faith –based remedies. And sometimes they turn to FDA-approved drugs.
The Cochrane Collaborative has reviewed the scientific literature on the use of cholinesterase inhibitors (like donepezil) in mild dementia, and has found:
There is no evidence to support the use of donepezil for patients with mild cognitive impairment (MCI). The putative benefits are minor, short lived and associated with significant side effects. http://www.cochrane.org/reviews/en/ab006104.html
So how did this drug get approved? Well, there do seem to be some small improvements (of dubious clinical significance in my opinion) in measures of cognitive impairment in patients with Alzheimer’s dementia in particular. http://www.cochrane.org/reviews/en/ab005593.html
AHRQ states:
The evidence is mixed, however, about the effects of cholinesterase inhibitors on functional measures such as instrumental activities of daily living (i.e., ability to use the telephone, mode of transportation, responsibility for medication, and ability to handle finances). In general, the studies show little or no effect on functional decline after 6 months of treatment and a small but statistically significant difference from placebo after 12 months of treatment.
Research has found no clinically important differences between people taking cholinesterase inhibitors and those taking placebo in the development of behavioral and psychological symptoms… Studies rarely addressed other important health outcomes such as utilization of health care services, injuries, and caregiver burden.
http://www.ahrq.gov/clinic/3rduspstf/dementia/dementsum.htm
Pfizer’s press release (when they received FDA approval to market Aricept in 1996) noted:
Alzheimer’s disease is a family tragedy. ARICEPT will benefit patients and families alike by improving or maintaining patient function, which in turn may help ease the burden for caregivers and help maintain personal dignity… “ARICEPT represents a significant step forward in addressing the therapeutic needs of the Alzheimer’s disease community…This therapy will help to change the approach to the management of Alzheimer’s disease.
http://www.pslgroup.com/dg/e2aa.htm
Global sales of Aricept were approximately $1.1 billion for 2008 alone.
Me-too cholinesterase inhibitors have seen similar global profits, with sales of namenda at about $1 billion as well in 2008. http://www.businessweek.com/magazine/content/04_14/b3877629_mz073.htm All this while the AHRQ can find no clinically relevant difference between the drugs in this class, and the effects they have are small and short lived.
There are pharmaceutical innovations that have changed the course of history (imagine where we’d be without the polio or smallpox vaccines), while others leverage the tiniest statistically-significant effects to drive global drug empires driven by public feelings of helplessness in the face of currently incurable diseases.
It’s no wonder that the public has a mistrust of pharma – their marketing engines drive sales of drugs that have vastly different clinical value. That means it’s up to physicians and scientists to tease out the legitimate enthusiasm from the marketing hype. And judging from all the patients with mild dementia that I see on cholinesterase inhibitors, I give us a failing grade.

I distinctly remember the day I attended a “drug lunch” (as a PM&R resident in New York City) to learn about the value of donepezil (Aricept) for the treatment of dementia. I was surprised by the drug’s lack of efficacy – the graph displayed in the PowerPoint show demonstrated a 2-point improvement on the Mini Mental State Exam (MMSE), an effect that began after 6 months of donepezil use, and persisted for only 6 months after that. A 2-point difference on the MMSE has no clinical relevance of which I’m aware. The drug’s common side effects include: nausea, vomiting, diarrhea, loss of appetite, tiredness, drowsiness, trouble sleeping, or muscle cramps. That day I realized that the risk-benefit profile did not support its use.

Nonetheless, I was perplexed by the number of patients who came to the hospital already on the medication. Over and over again I heard the same story: “Mom is becoming forgetful so our doctor started her on this medication to help her memory.”

When I asked the family if they thought the medicine helped, the response was equally predictable: a shrug and then “What else can we do?”

Here we have a classic example of a medical problem with no satisfactory treatment or cure – and a desperate desire on the part of patients and family members to do something – anything – about it. Many times people in these predicaments turn to alternative medicines, herbal supplements and faith –based remedies. And sometimes they turn to FDA-approved drugs.

Donepezil is approved for the treatment of mild to moderate dementia of the Alzheimers type. Although the benefits are modest, the drug can delay progression of symptoms for about six months.

Mild cognitive impairment (MCI) is mild memory difficulty severe enough to be noticed but not to affect activities of daily living. About 20% of people over 70 have MCI, and they are more likely to go on to develop dementia, even though many patients with MCI do not progress. Although donepezil is not indicated for MCI it is commonly used to treat it.  The Cochrane Collaboration has reviewed the scientific literature on the use of donepezil in mild cognitive impairment, and has found:

There is no evidence to support the use of donepezil for patients with mild cognitive impairment (MCI). The putative benefits are minor, short lived and associated with significant side effects.

So how did this drug get approved? Well, there do seem to be some small improvements (of dubious clinical significance in my opinion) in measures of cognitive impairment in patients with Alzheimer’s dementia in particular.

The Agency for Healthcare Research and Quality (AHRQ) states:

The evidence is mixed, however, about the effects of cholinesterase inhibitors on functional measures such as instrumental activities of daily living (i.e., ability to use the telephone, mode of transportation, responsibility for medication, and ability to handle finances). In general, the studies show little or no effect on functional decline after 6 months of treatment and a small but statistically significant difference from placebo after 12 months of treatment.

Research has found no clinically important differences between people taking cholinesterase inhibitors and those taking placebo in the development of behavioral and psychological symptoms… Studies rarely addressed other important health outcomes such as utilization of health care services, injuries, and caregiver burden.

Pfizer’s press release (when they received FDA approval to market Aricept in 1996) noted:

Alzheimer’s disease is a family tragedy. Aricept will benefit patients and families alike by improving or maintaining patient function, which in turn may help ease the burden for caregivers and help maintain personal dignity… Aricept represents a significant step forward in addressing the therapeutic needs of the Alzheimer’s disease community…This therapy will help to change the approach to the management of Alzheimer’s disease.

Global sales of Aricept were approximately $1.1 billion for 2008 alone.

Me-too cholinesterase inhibitors have seen similar global profits, with sales of rivastigmine at close to $1 billion in 2008. All this while the AHRQ can find no clinically relevant difference between the drugs in this class, and the effects they have are small and short lived.

There are pharmaceutical innovations that have changed the course of history (imagine where we’d be without the polio or smallpox vaccines), while others leverage the tiniest statistically-significant effects to drive global drug empires driven by public feelings of helplessness in the face of currently incurable diseases.

It’s no wonder that the public has a mistrust of Pharma – their marketing engines drive sales of drugs that have vastly different clinical value. That means it’s up to physicians and scientists to tease out the legitimate enthusiasm from the marketing hype. And judging from all the patients with mild dementia that I see on cholinesterase inhibitors, I think most of us deserve a failing grade.

Note: This article was updated on 01/25/2010

Posted in: Pharmaceuticals

Leave a Comment (41) ↓

41 thoughts on “Don’t Believe The Hype: Cholinesterase Inhibitors As A Treatment For Dementia

  1. lizkat says:

    Thank you for an excellent post! I suspected these drugs were next to worthless. Older people who are already on several drugs do not need the added side effects and interactions.

  2. Dr Aust says:

    In the UK, NICE (the national body that reviews drugs for the National Health Service in terms of effectiveness and value for money, and decideds if they should be recommended) did a full assessment of the cholinesterase inhibitors and decided they should only be prescribed under fairly restricted clinical circumstances (details here).

    The response of the drug companies (Eisai supported by Pfizer) was to sue through the courts to try and reverse the decision.

    The companies did not get the decision reversed, but did eventually win in court on one single count, the result of which which was that NICE had to release a full working version of its cost-benefit model. Full story here.

    Of course, no such transparency applies to the decisions the companies make as to how to price the drugs; their pricing models and strategies remain wholly opaque, on the grounds of commercial sensitivity.

    One of the more depressing aspects of the whole depressing business was the sight of respected patient (and research) charity the Alzheimer’s Society lined up squarely alongside Eisai and Pfizer and shouting for “every patient’s right” to have the drugs.

  3. delaneypa says:

    Val, I agree that Aricept is not a very good medicine overall, and we all wish there were something, anything better. But memory is only one of the problems with dementia, and the only thing measured by the MMSE. Far more important than memory impairment include loss of function, demands on caregivers, behavioral disturbances, placement in a nursing home, and ultimately death.

    While the effect on MMSE may be small, even small effects may be clincally meaningful in these other areas. Postponing nursing home placement (e.g. Lopez OL, et al. J Neurol Neurosurg Psychiatry. 2002;72:310-314) for say 6 months, would give the patient much more time with the family at home, AND save everyone thousands of dollars.

    One study showed that Aricept compared to placebo gives caretakers one less hour during which they have to directly monitor the patient (Feldman H, et al. J Am Geriatr Soc. 2003;51:737-744). As a caregiver that would be very valuable time!

    So an important discussion to have with patients and their families is what are the treatment goals and expectations. Yes Aricept is oversold (what drug isn’t?), but it has a place.

  4. marilynmann says:

    I agree. My mother’s internist wanted to put her on Aricept. I researched the drug, and said thanks but no thanks.

  5. Bentham says:

    First things first. You should correct the paragraph:

    “Me-too cholinesterase inhibitors have seen similar global profits, with sales of Namenda at about $1 billion in 2008. All this while the AHRQ can find no clinically relevant difference between the drugs in this class, and the effects they have are small and short lived.”

    Namenda (memantine) is not an acetylcholinesterase inhibitor at all. It is in an entirely different class. It is an NMDA receptor blocker. It is hard to for me to think of a charitable explanation for this blunder.

    The Folstein Mini Mental State Exam is a terrible test. It is only used because of clinical inertia. The sensitive portion of the test is the three item recall portion. In fact, there is good study evidence that you can combine a three item recall exercise with a clock drawing exercise (together called a Mini-Cog, designed by Soo Borson) and do just as well as a Folstein in considerably less time. My personal dementia screen in office combines the three item recall, clock drawing, one minute animal naming, and 5 item geriatric depression screen. This gives you a great deal more information that a Folstein in less time and with less patient irritation. (People seem to hate taking a MMSE.) If you want a longer evaluation tool, try a Saint Louis University Mental Status Examination. (Although I recommend changing the instructions for the clock drawing to “eleven-ten” rather than “ten minutes to eleven o’clock”. This was a lapse in their final draft of the test. Choosing 11:10 as a time assesses retention of knowledge of how a clock works and ability to translate that knowledge. An early clock-drawing error when you ask someone to draw “eleven-ten” is to have one hand pointing to 11 and another to 10. You will miss this error if you ask for “ten minutes to eleven o’clock”.

    The assertion that 2 points on an MMSE has no clinical importance is not true without significant restriction placed on the statement. If I did MMSEs anymore (which I don’t) and someone had a 28/30 and their 2 items missed were in three item recall, then I would probe further. This is only 2 points less than 30, which would probably end my assessment unless there were patient or family concerns that merited further evaluation. With 28/30 and missing two recall items (or perhaps one recall and the visuospatial point), you should have a fair suspicion of at least mild cognitive impairment if not dementia (especially in a highly-educated patient). You should initiate further questioning about activities of daily living as well as do further evaluation with another memory assessment tool. (Saint Louis or perhaps Montreal Cognitive Assessment.) Lab and imaging workup wouldn’t be unreasonable. Repeat cognitive testing would make sense and possibly a neuropsych referral if the patient repeated that level of deficit.

    In regards to donepazil and mild cognitive impairment, the NEJM 2005 study
    http://content.nejm.org/cgi/content/full/352/23/2379
    is negative if you ask the question, “Does donepezil prevent conversion of MCI to AD?” It is positive if you ask the question, “Does donepezil slow the conversion of MCI to AD?” There is also subgroup analysis that suggests APO E4 genotype patients are particularly likely to benefit.

    You are choosing to equate mild dementia and mild cognitive impairment. This is another error. These are not the same thing. It might be helpful in a family meeting to refer to mild cognitive impairment as a mild form of dementia, but it is professionally imprecise. (Actually, that is gentle. It is simply and straightforwardly wrong.) Mild cognitive impairment has a deficit in one+ cognitive area (most commonly amnestic). Dementia requires a deficit in memory plus another area AND a a detrimental effect on activities of daily living. Without the ADL (or IADL) impairment, you can’t have dementia.

    The side effect profile of AChE-Is is not particularly onerous in comparison to other pharmaceuticals we use. A decision to start one should always be done in context of a family conference if possible, with tempered expectations about the small amount of benefit that can be hoped for. The decision to start such a drug is also made in context of discussion of other interventions that may help. Physical and mental exercise and increased socialization, for instance.

    In general, my suggestion for most patients without contraindications IS to start such a medication. It isn’t a strong recommendation, and I don’t press.

    Besides being a geriatrician, I’m a hospice doc. This class of drugs is one of the most common that I recommend stopping. I do not consider this to be an inconsistent view at all.

    You speak about having attended a drug rep dinner as a rehab resident. I wouldn’t know about attending drug rep dinners as a resident. I never did. I considered it then and still consider it now to be unethical.

  6. Bentham says:

    Listening to a drug company sponsored talk and then making up your mind that a class of pharmaceuticals (the members of which you cannot even successfully identify) does not satisfy risk/benefit analysis is perhaps not a terribly Science Based Medicine approach.

    Perhaps if you had spoken first to someone in the actual specialization area in question, you might have found out that most of us don’t actually give a flying fig about someone’s change in Folstein score.

    We care about their change in functional status. There is reasonable evidence that donepezil delays loss of functional status. This is not a robust effect, only in the neighborhood of half a year on average. Half a year delay in the disabling effects of a disease that is now considered to have a clinical course of about 5-7 years is not trivial, however.

    This is not a well written article and should be removed.

    Before anyone starts to accuse me of being a Big Pharma shill, let me say that I don’t take gifts from pharmaceutical companies. I think they should be banned entirely (as should direct-to-consumer advertising). There is no doubt that multiple companies have exaggerated (to the point of lying) the effects of ACHEIs to both patients and doctors. They should continue to be sanctioned for doing so.

  7. Bentham says:

    Asking a patient’s family if they think that donepezil is helping is not a useful way to assess efficacy. Very few patients are actually going to get better at all with the drug. If a drug effect is only to slow the progression of disease (in about 80% of people who take it), how can a family member be expected to comment? I’ve seen a few patients who seem to have benefited clearly after starting donepezil, but how am I to separate out the effect from this drug rather than the effect of the simplification of the rest of their medication regimen or the fact that they’ve started attending adult day care or initiated an exercise program? Rather than anecdote, one should probably look to the study data. Once again, pretty queer for an SBM post.

    I’ve got a big census and have to go to bed. If there are any responses to my rants, I’m not ignoring them. I’ll be back tomorrow night to follow up.

  8. windriven says:

    “In general, my suggestion for most patients without contraindications IS to start such a medication. … This class of drugs is one of the most common that I recommend stopping. I do not consider this to be an inconsistent view at all.”

    While you do not consider yours to be an inconsistent view, it would seem to be inconsistent to a casual observer. Might you expand on your thinking here?

    “Asking a patient’s family if they think that donepezil is helping is not a useful way to assess efficacy.”

    Asking a patient’s family would be inappropriate as the sole criterion for rendering judgment but I don’t think that was Dr. Jones’ point. The family is likely to spend more time than anyone with the patient and has the best baseline against which to measure current condition – at least in a social sense. How can asking for family observations fail to be useful?

  9. nock says:

    Yeah, this article left a bad taste in my mouth. I agree that Namenda should not be lumped in with AChE inhibitors, since, well, they’re NOT AChE inhibitors. Plus, from what I remember, Namenda can result in significant cognitive improvements in very impaired patients (those with MMSE <15). This is Science-Based Medicine blog, please uphold the standards that are implied when declaring yourself Science-Based.

  10. Bentham says:

    Looks like I can’t get to sleep yet after all.
    First of all, I’d like to apologize for the tone of my replies.
    I was certainly pretty angry, but I ought to have edited more. There was no reason to be a boor.

    @windriven
    I often recommend starting them when I’m wearing my memory consultant hat.
    I often recommend stopping them when I’m wearing my palliative consultant hat.
    I am seeing patients at different points in their disease trajectory. There is some evidence (not yet completely convincing to me) that the effect of cholinesterases is finite. As a geriatrician doing a memory consult, I’m usually seeing people at the beginning or middle of their course. There is still a significant amount of function left to try to preserve as long as possible. As a hospice doc, I’m usually seeing patients whose functional abilities are currently very diminished. For a patient who has a single digit number of intelligible words per day, who is asleep most of the day, who is bedbound or nearly so, who is incontinent of bowel/bladder/both, who requires assistance with eating (or has largely stopped eating), the small effect of these drugs probably doesn’t justify their use. I’m looking at that point in time to simplify the regimen to medications that are likely to provide for patient comfort, and these ain’t them.

    @windriven point two:
    I am absolutely going to be asking for updates on functional decline. That is not the problem I have with the original author’s post. The claim certainly seemed to be that we could assert lack of efficacy based on lack of perceived efficacy from the point of view of the family. The problem is that the expected natural course of this disease is a progressive decline in memory and function. The expected course of the disease with drug treatment is a slightly slower progressive decline in memory and function. This includes some people with a more robust response, some with a less robust response, and some with no response. I do not believe that I would personally have a great success rate at trying to guess whether a patient was on donepezil or placebo even with serial exams over a year. A family cannot possibly be able to do so, especially considering that their experience with demented patients is likely to be quite limited. People have different velocities of disease progression as well. Say someone went from 90% functional to 50% functional (work with me on this abstraction) in 4 years. If they are on drug, I cannot tell if they were someone who would have been at 40% function without drug or if they are one of the 20% of people who don’t have any benefit at all from drug. If you ask me whether or not drug has helped this person, I cannot tell you. How on earth can the family tell you? The only way that we really can tell that this drug works is that there is a detectable difference in the means of several different measurements (including clinically relevant functional ability) in treatment group compared to placebo (and that the drug has sufficient biological plausibility such that a p-value of 0.05 is a reasonable cutoff and that studies are not guilty of failing to control for multiple comparisons).

  11. Acetylcholinesterase inhibitors are an interesting example for SBM really.

    I’ve done a bit of geriatrics here and there – Bentham feel free to completely correct me if I’m talking out of my league here – and the geriatricians I’ve worked with have generally said that some patients respond really well to this class of drugs, and some patients don’t really respond at all. They say they’ve seen people on a years long path of decline suddenly halt, and even make significant gains on the drug, and lots of other people just complain of nausea but show no improvement, or slowing of decline.

    Dementia seems to be one of those conditions with a pretty variable progression to begin with, and on top of that, it has a wide variety of possible etiologies, many of which we have limited understanding of the underlying pathologic processes. All of that would lead me to be completely unsurprised if we eventually find that there is just a class of patients for whom this is a great drug, and we just don’t know how to identify that group yet.

    That being said, the evidence indicating their use broadly across populations is weak, and I think that’s something which should be discussed with patients and their families individually like we would any other therapy for which we have limited or equivocal evidence. We have similar conversations about PSA tests for aging men for example.

  12. BillyJoe says:

    Nobody mentioned the cost:

    $5 per day in Canada, slightly less in USA, and slightly more in Australia.

    The question is: does the benefit justify the cost?

  13. Bentham says:

    @whitecoattales
    I’ve seen some people that improve, but I think it’s important to realize that this may just be temporally related to initiation of drug and not necessarily caused by it. For amnestic mild cognitive impairment (which is not the same thing as mild dementia–no matter what the original author believes), the expected natural course is quite variable. Some improve and never develop dementia. Some stay where they are at and likewise never develop dementia. Others progress to dementia. Something like 15% risk for conversion in a year. It would be nice to be able to determine which subgroup for whom MCI is actually prodromal Alzheimer’s. APO E4 might be a start, but I don’t know anybody who is doing it in regular practice. When an MCI patient improves after I have started them on donepezil, I have to be mindful that it might have happened anyway. It might also have more to do with the other drugs that I stopped or the nonpharmacologic interventions we started. Donepezil is off label for MCI. I only start it after a detailed discussion and feedback about what the patient wants and believes. The last MCI guy I started on Aricept is a competitive chess player for whom even a small cognitive deficit is likely to be felt, even though it won’t affect his self-care abilities. He would likely have wanted to stay on the drug even with mild GI side effects (which he didn’t have).

    The comparison with PSAs is a good benchmark. I don’t do primary geriatrics anymore. When I did, my advice on PSAs had slid a little bit as more data came in. In dementia, I’ll generally recommend drug but not push. In MCI, I’ll generally discuss drug as an option but remain neutral about it unless the person seems to be a good candidate due to their own goals and beliefs. For average risk men and PSAs, I was bringing up the subject and gently advising against. If, after discussion, it was still something they wanted, then I was certainly ordering it.

    @BillyJoe
    The cost was a key element in the previous British cost-benefit analyses that previously had them stopping acytelcholinesterase inhibitors upon admission to nursing home. It wasn’t that the drug didn’t work, it was just that the effect was small and the cost was relatively large. The cost is even higher when you add a second agent. It is quite normal to try both an AChE-I and a NMDA receptor antagonist for a moderately demented patient.

  14. tahoe69 says:

    This topic reminds me of a Chicago Tribune “expose” of how some powerful psychotropic medications are/were being administered to nursing home residents in Illinois withOUT consent and withOUT valid
    psychiatric diagnoses.
    In this piece Dr. Graham a FDA scientist said this to congress two years ago, he estimated that there are thousands of nursing home resident
    fatalities annually over antipsychotic drugs being administered unnecessarily and to patients who are NOT diagnosed with any mental illness.
    Researcher Christie Teigland is looking at data on some 275,000 nursing home residents who have been diagnosed with dementia. Teigland has found that those patients taking psychotropic drugs were likelier to fall
    or deteriorate versus other patients and when taken off the medications,
    tended to improve.

  15. pmoran says:

    I think Val is correct in her wariness concerning drugs like this i.e. with evidence of rather minor benefits for common, otherwise incurable chronic illnesses, where drug companies can rake in billions of dollars without contributing greatly, or possibly even at all, to the welfare of mankind. The companies will not care about the cost/risk/benefit considerations, and we doctors will collude, through the heavy pressures upon us to “do something!” for these folk

  16. Val Jones says:

    Thanks for the comments. I corrected the error in which I typed Namenda instead of rivastigmine.

    I think it’s reasonable to ask the family (or caregivers) about their observations regarding the patient’s behavior and symptoms. If they are unable to discern any appreciable improvement (or even absence of worsening) then one has to wonder how much of an effect on function and ADLs these drugs are really having.

    Dr. Aust – In this case I would agree with NICE’s assessment. Thanks for link.

    Bentham – My point was exactly that: like you, I don’t “give a flying fig” about 2 points on the Folstein score. And it’s alarming that this score change was purported to be proof of clinical efficacy.

    My point about donepezil is that it appears to be highly over-prescribed, especially for off-label uses (its only approved indication is mild to moderate Alzheimer’s disease). Both Cochrane and AHRQ have found little to commend this class of drugs, even for their one indicated use. Yet we are spending billions every year on them. Closer scrutiny of the risk/benefit analysis of cholinesterase inhibitors is warranted. It is clear that many patients are currently taking these medicines without any clinical benefit.

  17. BillyJoe says:

    Considering their small benefit and enormous cost, it was a surprise in Australia when the dementia drugs were placed on the PBS to receive government subsidy. The main reason this happened was that groups such as “Alzheimers Australia”, which is a patient self-help group, made a big public push for it to happen. It may be no coincidence that both Pfizer and Jansson-Cilag are sponsors of Alzheimers Australia. In recent times drug companies have actually been the instigators of self help groups (eg “Prostate Cancer Foundation of Australia”) who then lobby for media space to publicise the disease and its management.

  18. Bentham says:

    @Val Jones.

    It is good that you corrected your mistake about the mechanism of action of memantine.

    It is bad that you introduced another error now. You have just stated multiple times that the only approved indication is mild to moderate Alzheimer’s disease. You are just digging yourself in deeper here. Aricept has been FDA approved for severe Alzheimer’s dementia for over three years now. This is clearly outside your area of expertise. You’ve done other good work on this site, but this is not good work. Please just take it down.

    You have brought up the Folstein improvement again. Perhaps you did not know what studies on dementia include MMSEs. For drugs to get approval by the FDA for use in treating AD, they are required to show improvement on a cognitive scale. The Folstein is by far the most widely used scale. The FDA set the bar on this one. It isn’t a matter of the companies trying to pull a fast one (not that they wouldn’t do so). If the FDA says you have to show an improvement on a cognitive measure, and says that they don’t care about a subjective caregiver report questionnaire (which many practicing specialists care much more about), then trials are going to be tailored to the measures that the regulatory agency wants.

  19. Bentham says:

    Here is a paragraph-by-paragraph review of the original SBM post:

    @paragraph 1: Author describes her experience at a drug company sponsored meal. Some of us (too few to my mind) consider taking drug company gifts of any sort to be unethical. Reasonable data exists to suggest that information disseminated at such functions is not accurate. It is possible that the author was making a joke or engaging in hyperbole when she said that she realized that day the the risk-benefit profile did not support its use. Author also asserts that a 2 point MMSE improvement has no clinical relevance of which she is aware. The fact that she is not aware of it does not make it untrue. I gave an example of a situation where a two point MMSE difference would make a huge difference in the way an evaluation would proceed. A 30/30 is likely to stop the workup. A 28/30 (with 1/3 on recall) would mean full speed ahead (and already makes a diagnosis of MCI or mild AD probable).

    @paragraph 2/3: Based on the natural course of the disease and the expected benefit of the drug, surveying the families is not a reasonable way to assess for efficacy of this drug in this disease.

    @paragraph 4: This certainly is a disease without cure. It is not a disease without treatment, however. There is literature data for multiple treatments for AD. The clinical experience of specialists is also that there is benefit. If you want to claim that these are anecdotes and that the plural of anecdote is not data, that is perfectly reasonable. But then please kindly remove all the anecdotes from your post.

    @paragraph 5: Author links to a Cochrane Collaboration review on mild cognitive impairment. She erroneously says that it is about mild dementia. This is not a minor error. This is a blunder. This admits to lack of qualification to discuss the topic authoritatively. One can certainly discuss topics outside their area of expertise and specialization (I do it all the time), but you have to be very careful about it. I believe that Dr. Novella’s practice is not focused on dementia, but he is still trained as a neurologist. Letting him take a peek at this before you posted it would have allowed correction of some of the areas. He very likely would have pointed out the mistake in drug class as well as the mistake in confusing mild dementia with mild cognitive impairment. I will discuss the actual Cochrane review in a follow-up post.

    @paragraph 6: “of dubious clinical significance in my opinion”. Different opinions hold different weight. In this case, you have shown that you do not have a deep professional understanding of this disease and the treatment options for it. Your opinion should not carry much weight. This is not even a situation where one has to rely on opinion. There really is data here. I realize that this still sounds like an attack. I cannot edit it any further and still retain meaning. By the way, in case you were wondering, I do feel significantly attacked (as a professional) by your post and follow up. You have called into question the way that I practice medicine (not calling me out personally of course). The fact that you do so while having far less knowledge than me (or even geriatrics fellows or internal medicine residents) is very, very galling.

    @paragraph 7: You are guilty of selective quoting. The AHRQ summary agrees that these studies found benefit in cognitive and global areas. They assert that the studies were well designed. The reason that all the studies focus on cognitive and global areas is that this is what the FDA asked for for approval. The studies often didn’t do functional assessment or behavioral assessments or perhaps they just tacked them on. This is a fault of the way that drugs are researched and approved in the US.

    @paragraph 8: Not much to say here. I’m certainly not a fan of drug company marketing. I certainly would place much greater restrictions on it than currently exist. (I might be a little too extreme in this area for most of the commenters and readers here.) Of note, the discussion/commentary sections of studies commissioned by drug companies often is quite different from those that are not. Similar magnitudes of effect are treated as monumental in one and of minor clinical significance in another.

    @paragraph 9/10: Statins also make a lot of money for their companies. This does not mean they are not effective. It is off topic for your post, but this might cause one to think about the way that the US awards patent protections and the way that they (largely don’t) regulate costs. Originally paragraph ten included the mistake which miscategorized memantine as an acetylcholinesterase inhibitor.
    Paragraph number 10 also asserts that “the AHRQ can find no clinically relevant difference between the drugs in the class”. Um, who cares? This is very common. Head-to-head trials are expensive gambles. What happens when you don’t find a difference or you even (heaven forfend) find that your drug is inferior? You have now spent a lot of money in order to provide your competitor with an advertising point. The current system encourages drug companies to carefully tailor studies in such a way that they can get FDA approval. At this point in time, they have very little incentive to continue to do studies. They now try to increase their market shares with aggressive advertising to both professionals and consumers.

    @last paragraph. You give me a failing grade for my clinical practice. I give you a failing grade for your clinical understanding and (so far) your reluctance to admit and correct most of your clear and egregious errors.

  20. Bentham says:

    Let’s actually do a little bit of science here.
    The cochrane review that Dr. Jones linked to was a review of mild cognitive impairment (despite her uncorrected erroneous belief and assertion that it is about mild dementia). I do not have a free link that I can share to the full review. I do have a link to one of the studies. Let’s review that study now.

    For those of you who will be bored by a technical journal article review, I’d recommend skipping to the end, where I’ll talk about clinical relevance of use of donepezil.

    Vitamin E and Donepezil for Treatment of Mild Cognitive Impairment. NEJM, June 9, 2005. Ronald C. Peterson, et al.
    http://nejm.highwire.org/cgi/content/full/352/23/2379

    Are the results of this study valid? Are the results of this study important? My own answers to these questions are yes. I’ll start with validity. I’ll largely follow the CEBM recommendations for validity/importance.

    I will be ignoring the Vitamin E group in the following analysis, treating the study as a straight RCT comparing donepezil to placebo and leaving out the vitamin E patients altogether.

    Randomized? Yes.

    Allocation concealed? Probably, but insufficiently explained in Study Design section. Could have described it better. (Few of us actually care, so authors often abridge here.) It would be nice if journals actually enforced inclusion of this information, as meta-analyses have suggested that studies with poor allocation concealment methods are slightly more likely to find treatment effect than similar studies with good allocation concealment methods.

    Were the groups similar at the start of the trial? Yes, by table 1

    Was follow-up of patients sufficiently long and complete? Three years follow up subjectively feels long enough. There is a concern about the dropout rate. 66 placebo dropouts and 92 donepezil dropouts. How did the authors deal with these dropouts? In short, the reevaluated the data with hypothetical assumption that double the rate of conversion to AD occurred in the donepezil dropouts compared to the completers. Even with the unfavorable assumption (which is what you really have to do to reassess honestly), there was still evidence for effect at 6 and 12 months. Lost the effect at 18 months which was there when not making this correction. The dropout rate for the study is still fairly high. The study would not have been considered strong enough to be published by ACP Journal Club because of it. They claim that there were no Table 1 differences between dropouts and completers. There were 230 total dropouts (including the vitamin E arm). 47 were for adverse events and 105 were for “withdrawal of consent”. There is insufficient explanation of the withdrawals. No tabular data for them is presented.

    Intention-to-treat? Yes

    Clinicians blinded? Yes

    Patients blinded? Yes

    Corrections made for multiple comparisons? (This is not one of the CEBM criteria.) Authors were very transparent about this. Sometimes they made these corrections and sometimes they didn’t. They explicitly told us when they were not doing so.

    Overall assessment of validity? Generally okay. Poor on dropouts. Failed to make a fundamental correction for multiple comparisons, but were transparent about this so that we can make it ourselves. As an ACP Journal Club article, I’d fail it. At a more normal standard of validity, I’d pass it.

    Is this study important?
    [I am first moving far away from the CEBM criteria. I think that this is an area that they overlook. As an upfront admission, I'm much more of a Bayesian than a frequentist. Those of you who are frequentists and can somehow feed and dress yourselves and figure out how to work a computer :) can skip ahead...]

    Is the intervention biologically plausible? Yes. There is evidence that amnestic mild cognitive impairment is frequently the prodromal phase of alzheimer’s disease. The yearly progression is considered to be in the neighborhood of 10-15% from MCI to AD. There is evidence that acetylcholinesterase inhibitors have a statistically significant effect on AD. There is a marked cholinergic deficit in AD patients. Donepezil increases the amount of acetylcholine in the cleft by inhibiting the breakdown esterase. This is sufficient biologic plausibility for me personally to believe that the pre-test probability for effect from donepezil in prodromal amnestic MCI is greater than 50%. Everyone is going to have to come up with their own pretest probabilities of course, which is one of the things that the frequentists can taunt me about. With this level of pretest probability, a p-value of 0.05 is reasonable in context of importance of the disease, lack of robust alternative therapies, and relative lack of known disabling side effects of the treatment.

    What is the magnitude of the treatment effect? None, if the question is “Does donepezil completely prevent MCI from progressing to AD?” Any clinician who is advising a patient or their family that this is the goal of treatment is wrong.

    If the question instead is “Does donepezil slow progression of MCI to AD?” then this study suggests the answer is yes.

    Use the one year data. 38/259 placebo and 16/253 donepezil patients converted to AD. The Control Event Rate (CER) was 38 divided by 259 or 14.7%. The Experimental Event Rate (EER) was 16 divided by 253 or 6.3%. Relative Risk Reduction (which is not terribly useful) is (CER-EER)/CER= 57.1% The much more useful Absolute Risk Reduction is CER-EER=8.4%. Number needed to treat is 1/ARR=11.9 (Would commonly be reported as NNT=12.) You would have to treat 12 people with amnestic mild cognitive impairment to prevent one of them from progressing to alzheimer’s disease at one year. You can see why this is not a strong recommendation for me.

    How precise is the estimate of the treatment effect? Here we will have to simply use the author’s published data. They claim a hazard ratio at one year of 0.42 with a 95% CI of 0.24 to 0.76. They report this as a p-value of 0.004. They explicitly admit that this is not corrected for multiple comparisons. If you want to make a correction for multiple comparison, a quick and dirty way to do this would be to do a Bonferroni correction. Since they gave hazard ratio data at three different points, 1, 2, and 3 years, we should not be using 0.05 as our cutoff but rather 0.05/3 or 0.0167 (which they still beat).

    ————————-

    Final verdict. Reasonably-designed study. Did not show that donepezil could stop progression of amnestic MCI to AD. (With a retrospectoscope, I can say that this is not a surprise at all. It can’t stop progression of AD either.) It did provide data to suggest that it can slow progression. Given that it is biologically plausible and has actually got a relatively benign side effect profile, it is reasonable to use. Side effects are usually GI-related, they are usually mild, and THEY ARE REVERSIBLE. If you do not tolerate the drug, you can stop it. When I start the drug, I use a step-up approach (which reduces side effects) and I schedule a 3 or 4 week follow up to assess tolerability.

    Something in the neighborhood of 80% of MCI patients will progress to AD. The rate is about 15% per year. By the result of this study, number needed to treat for slowing progression at one year is only 12. If I treat 12 MCI patients with an AChE-I, on average, one fewer person will have developed Alzheimer’s disease at one year.

    Personally, I’d sign up for that deal in a heartbeat. (If I were 30 years older and had MCI.)

  21. David Gorski says:

    Please just take it down.

    As managing editor, I’m very curious why you have repeatedly demanded that this post be taken down. Such a demand strikes me as very odd indeed. It is actually a pretty rare thing for one of our commenters to demand so vociferously that a post be deleted, no matter how much he might disagree with the post or error-ridden he thinks it. To give you an idea of how uncommon it is, the last time it happened was when Trevor Marshall demanded that Dr. Crislip’s post about his protocol be removed; he even threatened legal action in e-mails to us.

    http://www.sciencebasedmedicine.org/?p=563
    http://www.sciencebasedmedicine.org/?p=681

    That was really the only time I can recall in our two year history someone insisting as strongly as you that a post on SBM be removed.

    We at SBM are, of course, committed to scientific accuracy; however we are not all experts in every subject. Sometimes we will get something wrong and will have to correct it. When that happens, it is important to remember that this is a blog, not peer-reviewed literature. (at least not pre-posting). Our “peer review” occurs after the post is published and it takes place in the comments, as it has for this post, and we do ask our bloggers to correct factual errors when they are pointed out and verified.

    In this process, you have had more than ample opportunity (indeed, you’ve had free rein) to voice your objections in comments whose length rivals that of a full post by me or Kim Atwood (the two most logorrheic SBM bloggers), and Val has corrected the one clear error regarding memantine. Come to think of it, you’ve had far more chance to express your disagreements with this post than you would be likely to be given in any other venue that I can think of, including our tolerating your (at times) rather rude and insulting attitude. So why do you insist on this post being “taken down” when you have had such ample opportunity to criticize it in the comments? The rebuttals are there, appended to this post. You’ve made your point. Repeatedly.

    Finally, from my perspective nothing you’ve posted has really refuted the central points of Val’s post, namely that these drugs appear to benefit relatively few people, that their benefits are incredibly modest at best, that they are very expensive, and that they are heavily marketed by drug companies. As Val pointed out, Cochrane reviews don’t really support the widespread use of these drugs for MCI, and, as Dr. Aust pointed out, NICE rejected them as not being cost-effective (a reasonable conclusion, in my opinion). Even the study you just cited isn’t exactly what I would call a ringing endorsement of these drugs, even though it did show a statistically significant effect. In light of this research, it is not at all unreasonable to question, as Val did, whether their widespread use in MCI and Alzheimer’s disease represents the wisest use of scarce medical resources.

    You, of course, are perfectly free to disagree in the comments, as you have been doing. Free and open debate is how we roll at SBM. We correct factual errors when pointed out, but it would be an incredibly rare event for us to delete a post over complaints.

  22. BillyJoe says:

    Bentham,

    Thanks for you input into this thread :)

    “Aricept has been FDA approved for severe Alzheimer’s dementia for over three years now”.

    As a matter of interest, in Australia, Acetylcholine Esterase Inhibitors such as Aricept/Donepezil are indicated and approved under the Pharmaceutical Benefits Scheme for mild to moderate Alzheimers Disease, while NMDA Receptor Antagonists such as Ebixa/Memantine are indicated for moderate and severe Alzheimers Disease but are not covered by the Pharmaceutical Benefits Scheme.

    “Some of us (too few to my mind) consider taking drug company gifts of any sort to be unethical. Reasonable data exists to suggest that information disseminated at such functions is not accurate.”

    Glad to hear it. The most egregious result is that most GPs (in Australia at least) get all their CME points by attending Pharmaceutical Company “educational events” at which they are wined and dined to the tune of 300 million a year. It amazes me that this is allowed to continue.

    “The clinical experience of specialists is also that there is benefit.”

    I know you have a disclaimer there, but I think you should have avoided that sentence. Personal experience has no place, except merely as a starting point, in SBM. And that includes the personal clinical experience of experts.

    “They now try to increase their market shares with aggressive advertising to both professionals and consumers”.

    This is another thing I find difficult to understand: direct to consumer advertising. This doesn’t happen in Australia but, for some reason, they are allowed to advertise diseases treated by their drugs and then tell the consumer to “ask your doctor” about the treatment.

    “You would have to treat 12 people with amnestic mild cognitive impairment to prevent one of them from progressing to alzheimer’s disease at one year. You can see why this is not a strong recommendation for me.”

    For me that is an understatement!
    What you mean, and correct me if I’m wrong because you seem to have a better grasp of this than I do, is that without treatment about 12 patients out of 100 (10-15%) with MCI will progress to AD each year. With treatment, that number is reduced to 11.
    And I hope my maths are correct but, in financial terms, that comes to an outlay $5 X 365 X 12 = $22,000 to reduce the number of patients progressing from MCI to AD each year from 12 to 11 per hundred patients. And that doesn’t include the cost of the evaluation, paperwork, prescription etc.

    “If I treat 12 MCI patients with an AChE-I, on average, one fewer person will have developed Alzheimer’s disease at one year”.

    Have I missed something or have you mis-stated the conclusion?
    Of your 12 patients with MCI, 1.4 will progress to AD each year without treatment”. Are you really intending to say that with treatment you will prevent 1 on these from progressing? If not, I think your statistic is misleading.

    regards,
    BillyJoe

  23. weing says:

    All my experience with these meds is anecdotal. I am not impressed by them. Maybe they slow down the progression of the disease, but that is so hard to determine. I have seen dramatic improvement only once and that was on rivastigmine. One patient, on galantamine, has been functioning on her own for several years. But is it the drug? I have no way of knowing.

  24. BillyJoe says:

    Bentham,

    “If I treat 12 MCI patients with an AChE-I, on average, one fewer person will have developed Alzheimer’s disease at one year”.

    I think I got it wrong.
    Re-reading it, in that trial, without treatment approximately 15 out of a 100 with MCI progressed to AD in the first year. With treatment the figure was about 6.
    I suppose the clearest way to state this is to say: “Out of 100 patients with MCI, the number that will progress to AD will be reduced from 15 to 6 each year with treatment.

    David Gorski,

    “including our tolerating your (at times) rather rude and insulting attitude”.

    You should have a look at the “Is breech vaginal delivery safe?” thread!!!

    Weing,

    “I have no way of knowing”.

    You are right, personal experience just doesn’t cut it. ;)

  25. David Gorski says:

    You should have a look at the “Is breech vaginal delivery safe?” thread!!!

    True that. The torrent of nastiness in that thread was beyond my capacity to staunch, I’m afraid, but it was as nothing compared to the circumcision post. In any case, the reason Bentham got on my nerves was through his repeated demands that Val (or we editors) delete this post. So I asked him why. We’ll see if he answers.

  26. Bentham says:

    @ David Gorski

    Of course I’ll answer.

    You say “As managing editor, I’m very curious why you have repeatedly demanded that this post be taken down. Such a demand strikes me as very odd indeed. It is actually a pretty rare thing for one of our commenters to demand so vociferously that a post be deleted, no matter how much he might disagree with the post or error-ridden he thinks it.”

    It seems that I have asked twice for it to be taken down:

    “This is not a well written article and should be removed.”

    “You’ve done other good work on this site, but this is not good work. Please just take it down.”

    I suppose that since I said something twice, this technically means that I said something repeatedly.

    You assert that there was only one clear error in the author’s posts and that it was corrected. This is not so. There are multiple other errors, at least two of which are not disputable. Original author asserted that mild cognitive impairment and mild dementia are the same thing. They are not. Original author asserted that the only FDA approved indication for donepezil is mild-to-moderate AD. It is not. It has been approved for severe AD since 2006. These are clear errors.

    You can do whatever you want, of course. It’s your sandbox. I’ll just move along. This is not the place that I expected it to be based upon the names I see represented.

  27. Peter Lipson says:

    B, I think you have a fundamental misunderstanding of the medium. Blogs are interactive. The quality of medical writing on this blog is some of the best in the blogosphere, but is very different from, say, a journal article.

    But one of the advantages is that readers—such as you—can point out putative errors, influence the final draft, and have your ideas appended to the original.

    It’s unique; it’s new. It’s more like a conference than an article.

    There seems to be a forest/trees issue here. You are rather concerned about the minutiae (which are still important) but you seem to be over-extending your critique. The criticism, “You got some things wrong, therefore the conclusion is wrong” is not a convincing argument unless the things that were wrong either logically invalidate the argument or are so fundamental to the point that the argument is destroyed.

    It is good to point out factual flaws, but it is not good to lose sight of the original argument: that pretty much all drugs for dementia are disappointing, not terribly useful, full of side effects, over sold, etc.

  28. David Gorski says:

    I suppose that since I said something twice, this technically means that I said something repeatedly.

    Very well. I intentionally did not mention this in my previous response to you, but you leave me little choice. What our readers don’t know is that you also e-mailed Steve strongly suggesting that we “consider” taking Val’s post down. That makes three times, two of which were pretty insistent. I didn’t want to mention your e-mail to us and intentionally tried not to. But if you’re going to play the numbers game to try to deflect this particular question of mine, then I feel obligated to point it out, as writing an e-mail to the editors goes far beyond just commenting in my book. Moreover, the reason your requests to delete Val’s post stand out in my mind is because even one such insistent request is unusual. We’ve had readers criticize various members of the SBM team, sometimes very harshly; we’ve even had readers say that an SBM blogger should be booted from the team as being “unworthy” of SBM. However, it’s pretty uncommon for someone to demand that a specific post be deleted. Such requests stand out.

    Confirmation bias? Maybe, maybe not. But, as I said, I can only remember one example prior to yours of such a demand stated so emphatically. I also note that you have not answered my question.

    You assert that there was only one clear error in the author’s posts and that it was corrected. This is not so. There are multiple other errors, at least two of which are not disputable. Original author asserted that mild cognitive impairment and mild dementia are the same thing. They are not. Original author asserted that the only FDA approved indication for donepezil is mild-to-moderate AD. It is not. It has been approved for severe AD since 2006. These are clear errors.

    And, as both Peter pointed out, none of these invalidates Val’s point, which is that these drugs don’t help very many people, are expensive, have side effects, and are heavily marketed by pharma. There is no error in this post that is “fatal” to her message, and Val will correct any factual errors, as we always strive to do here at SBM.

    Peter said it very well. You strike me as not understanding the nature of blogs and of trying to use relative minutiae to discredit the main points about these drugs that Val was trying to make. Perhaps as an expert you find the minutiae to be very important. And so they are in many contexts; however, as important as you seem to think they are in this context, even in the worst case scenario where each and every error you pointed out is indisputably an error, none of these factual errors repudiates any of the major points of Val’s post. There’s where you seem to be trying to discredit the point of the post with what are, in the context of the overall points Val was arguing, relatively minor errors. Moreover, I’ll repeat that nothing you have argued or presented refutes Val’s main points about how these drugs don’t help very many people, are expensive, have side effects, and are heavily marketed by pharma. The one study that you did present was far from a ringing endorsement of these drugs.

    You can do whatever you want, of course. It’s your sandbox. I’ll just move along. This is not the place that I expected it to be based upon the names I see represented.

    That is your choice. However, by and large, I would put the quality of this blog up against the quality of any other medical blog anywhere. I also think that you have a rather unrealistically different expectation of what a blog is and how it functions than what is reality.

  29. Bentham says:

    @David Gorski

    I didn’t mention the email either. It didn’t seem appropriate. As long as you are mentioning it, however, you might have noted that the response was a confirmation from Dr. Novella that my concerns were valid, that a remedy would occur (but that removal was not the only remedy), and that at a minimum revisions were necessary.

    This is far from my first interaction with many of the people listed as staff. We’ve interacted at blogs (medical, scientific, political, and skeptical), conferences, and mailing lists. The list of names here is impressive. It really is. When I said that Dr. Jones had done good work in the past, I wasn’t blowing sunshine. Just because I hadn’t (I don’t believe) commented previously on this site under “Bentham” doesn’t mean that I haven’t been reading it.

    It is entirely possible that my hopes were unrealistically high here. Nobody here knows me under this current pseudonym (and I’m very small potatoes anyway). I still don’t really understand how it is possible to post the sorts of things people here post and use their own names. I used to post with my name, and am now on pseudonym number three (first two having been outed). I actually probably could be outed again based on what I’ve written here so far. I’ve received personal and professional threats previously, and as I say, I am very small potatoes compared to your staff. One of your editors is currently facing some unpleasantness at his other blog, and I couldn’t take that sort of drama.

    The issue is clearly a very emotional one for me. I’m still upset. I felt professionally attacked by the original piece. It is possible that I can’t actually communicate effectively (or even perhaps entirely rationally) on the issue because of this. I strenuously disagree that the errors in the original piece are trivial, but I seem to be in the minority here. Of note, in real life, I’m usually on the other side of this topic–usually warning that the data on AChE-Is do not support as robust of an effect as clinicians and family seem to expect.

    I’m well into logorrhea territory here :) , so I’ll wrap up. I’m not helping here right now. I’ll bow out. I’ll keep lurking. Perhaps I’ll post again in the future, maybe with pseudonym #4. If anyone wants to ask a question or continue a conversation with me, they can do so via jeremybentham AT email DOT com.

  30. EdinA2 says:

    Current controversy aside, I’d like to comment on the AChE inhibitors themselves. I am currently a resident and have limited experience with these medications. However, it is my understanding that these medications are not disease modifying agents, and will not slow the progression of Alzheimer’s Dementia. They will, perhaps, slow symptom progression by 6-12 months, although symptoms will eventually catch up when compared to patients not treated. The NMDA antagonists are similar, as they also slow symptom progression only, and do not affect the underlying disease. This is how I present the medications to my patients when we are making the decision to possibly try the medications.

  31. Zoe237 says:

    Thanks for this post, Dr. Jones. I will certainly be referring back to it should it become relevant in the future (knock on wood!).

    I think it’s *imperative* to correct factual errors should they become known, but I haven’t seen anything that questions the conclusion. I also really appreciate Dr. Bentham’s counterpoints. This is the sort of post that is truly helpful to those of us not pertaking in the woo, but not completely accepting of all mainstream medicine either. But I’m the annoying kind of patient who asks for a second opinion.

    I hope that medicine doesn’t evolve to the point where only highly specialized people (who may not have a lot of “perspective”) can evaluate data in their field. That’s the beauty of the blog. Of course, peer reviewed journals will continue to be my main source of information, but this format is really helpful too, especially as a place to start.

  32. David Gorski says:

    The issue is clearly a very emotional one for me. I’m still upset. I felt professionally attacked by the original piece.

    Why?

    No, seriously, why did you feel professionally attacked by the original piece? Even trying to imagine an emotional connection to the issue, I have a hard time seeing it.

  33. Val Jones says:

    I agree that I made an error in my comment response above. I said that Aricept is indicated for mild-moderate Alzheimer’s (which is true of other drugs in the class) but its indication has recently been expanded to include all stages of Alzheimer’s. This does not change my point that many patients are receiving Aricept for off-label uses, and that its clinical benefit for approved uses is small at best , while its clinical benefit for those without Alzheimer’s (especially “mild cognitive impairment”- as per Cochrane) appears to be close to zero.

    I agree that my final sentence was slightly provocative – because this is a blog and I wanted to spur discussion about the cost/benefit of this class of drugs. As skeptics we are quick to point out placebo-based hype (the fallacy of “miracle cure x”) but perhaps not as quick to discuss profit-driven exaggeration of some tiny, but measurable, “benefit.”

    Physicians and NPs may continue to fall prey to the temptation to over-prescribe if they don’t take a look at the real benefits behind the marketing enthusiasm. In my opinion, in the case of cholinesterase inhibitors, the hype is especially unwarranted.

  34. FYI – The original post has been edited to correct any factual errors, and a notice to that effect was placed.

  35. Bentham says:

    @Val Jones
    “I said that Aricept is indicated for mild-moderate Alzheimer’s (which is true of other drugs in the class) but its indication has recently been expanded to include all stages of Alzheimer’s. This does not change my point that many patients are receiving Aricept for off-label uses, and that its clinical benefit for approved uses is small at best , while its clinical benefit for those without Alzheimer’s (especially “mild dementia “- as per Cochrane) appears to be close to zero.”

    The expansion of indication for Aricept to all stages of Alzheimer’s dementia was not recent. It happened in 2006. The authors of the Cochrane review that you cited included it in their summary: “Donepezil (Aricept, E2020), a cholinesterase inhibitor, has been shown to benefit all severities of AD including mild and it is reasonable to investigate its efficacy for those with MCI.”

    Are you asserting that the off-label prescription of medications is improper? Or are you asserting that the fact that a medication is not on label means that its use is not evidence-based?

    Are you still arguing that mild cognitive impairment and mild dementia are the same thing? It looks like you are suggesting that the Cochrane authors are saying this. Could you please point out the spot where they said this?

    If you no longer consider mild cognitive impairment and mild dementia to be the same thing, what do you mean by your last sentence? By a straightforward reading, it looks like you are saying that those people who are prescribing Aricept for mild dementia are no better than those recommending alternative medicines and faith-based remedies.

  36. BillyJoe says:

    Bentham,

    “The authors of the Cochrane review that you cited included it in their summary: “Donepezil (Aricept, E2020), a cholinesterase inhibitor, has been shown to benefit all severities of AD including mild and it is reasonable to investigate its efficacy for those with MCI.””

    There is this Cochrane Review:
    Donepezil for mild cognitive impairment
    http://www.cochrane.org/reviews/en/ab006104.html

    Summary (a more complete version ;))
    “Donepezil (Aricept, E2020), a cholinesterase inhibitor, has been shown to benefit all severities of AD including mild and it is reasonable to investigate its efficacy for those with MCI. There is little evidence that donepezil improved cognitive function, and no evidence that donepezil delays progression to AD, but it was associated with significant side effects. There is no evidence to support the use of donepezil for patients with MCI.”

    Results:
    “In the first study…showed benefit associated with 10 mg/day donepezil compared with placebo at 24 weeks (MD 1.90, 95% CI 0.51 to 3.29, p=0.007), but four other measures of cognitive function did not.”
    “In the second study there was a significant difference between the number of patients diagnosed with AD or another dementia between the donepezil group and the placebo group in favour of donepezil after one year of treatment (16/253 donepezil, 38/259 placebo)…but no difference after 3 years of treatment (63/253 donepezil, 73/259 placebo).”
    “The third study assessed cognitive function but did not report the results.”

    Conclusion:
    “There are two included studies which reported results for cognitive function. One study demonstrated a modest treatment effect in cognitive function…but not for other outcomes assessing different domains of cognitive function.
    Donepezil was associated with significantly more adverse effects compared with placebo, mostly gastrointestinal. From the second study, there is no evidence that donepezil delays the onset of AD. There is no evidence to support the use of donepezil for patients with MCI. The putative benefits are minor, short lived and associated with significant side effects.”

    From one of your previosu posts:
    “Personally, I’d sign up for that deal in a heartbeat. (If I were 30 years older and had MCI.) ”

    Well, at this point, I’m wondering why you would do that based on the evidence presented.

    regards,
    BillyJoe

  37. yeahsurewhatever says:

    So how did this drug get approved?

    If you have the right keys, the FDA is a revolving door.

  38. Fifi says:

    This link isn’t directly on topic but it’s an interesting look at some suspect goings on linked to Dr. Lawrence Dubuske resignation from Harvard’s Brigham and Women’s hospital and how academic science gets corrupted (and the efforts being made to stave off corruption).

    http://carlatpsychiatry.blogspot.com/2010/01/dr-lawrence-dubuske-me-myself-and-irine.html

    The Carlat Psychiatry Report seems like it deserves a shout out in regards to SBM too :-)

    http://www.thecarlatreport.com/

  39. BillyJoe says:

    Fifi,

    We should not be surprised.
    Pharmaceutical companies are businesses and they have to sell to make profits. And it seems they use every stragedy known to man to do so. It is really up to the medical colleges and individual doctors to deal with this problem. Except for a minority of doctors, however, they don’t. In fact, most docs get all their CME points from drug company sponsored “educational” events and most docs continue to claim they can’t be influenced. :D
    In Australia, the $300 million spent by drug companies every year on CME activities for docs says otherwise.

  40. Fifi says:

    BillyJoe – I’m not the least surprised, both my parents are doctors so I don’t have an idealize image of medicine and research (be it idealized positively or negatively). And, hey, who doesn’t love the kitsch of having a Prozac pen or one shaped like a spine? Swag, every industry has it (the entertainment industry generally has much better swag though ;-)

    Individual doctors and researchers deal with these things differently. Some go as far as Dubuske, some stay ethical – but obviously it’s easier to stay ethical if universities and hospitals make it clear that unethical behavior isn’t acceptable and will be punished. There’s a reason why most pharmaceutical companies put more money into advertising, promotion and lobbying than they do into actual research (and why they’re pushing for public research to be diverted towards corporate ends).

    While I think it’s important to take on alt-med for the billion and one fallacies and corruptions, it’s equally important to SBM to be aware of the undue influence of corporations or ideologies that corrupt SBM and the ethical practice of medicine.

Comments are closed.