Cholinesterase inhibitors for patients with Alzheimer's disease: systematic review of randomised clinical trials
BMJ 2005; 331 doi: https://doi.org/10.1136/bmj.331.7512.321 (Published 04 August 2005) Cite this as: BMJ 2005;331:321All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Letters to the editor intend to discuss, broaden or question the
conclusions of articles published in the journal. They are part of
the search for improvement of knowledge. In some cases,
however, this objective is not reached.
On November 1 the BMJ published a letter from Prof. Förstl (“how to
do it properly!”) in which he seems to quote from an article
written by me in a German journal. As this German journal is not
available to most of the readers of the BMJ, I emphasize that Prof.
Förstl misquoted from this article. The incriminated list
he mentions is by no means an “advice not to use these drugs” in
the listed groups of AD patients. What is listed are the groups not
taken into account by several leading German psychiatrists when
arguing an undersupply of cholinesterase inhibitors up to 90%.
The arguments in this article concern questions of epidemiology
and health services provision, not clinical ones.
Deliberate misquoting is not a new technique to try to ridicule an
author. The aim seems to be to discredit the review on
cholinesterase inhibitors recently published by my colleagues and me. This review came to the conclusion that the quality of
the RCTs on these drugs is questionable and thus the validity of
the results is poor. Neither in this review nor in other articles was any
advice given on the use of these drugs in clinical settings. We
appreciate a rational debate on our position and on the
conclusions to be drawn from it. We will, of course, not join the level
of debate offered by Prof. Förstl.
Prof. Dr. Hendrik van den Bussche
(bussche@uke.uni-hamburg.de)
Competing interests:
None declared
Competing interests: No competing interests
The work of the Hamburg-group has been criticised for being critical
without offering constructive recommendations. The senior author (1) has
now summarized his thoughts for the German readership on
how to use acetylcholinesterase-inhibitors properly. He advises us not to
use
them in:
• Patients suffering from relevant side-effects
• Patients not eligible for treatment because of foreseeable side-
effects
• Patients refusing treatment
• “Burned-out cases” (“Austherapierte”) showing deterioration after
an initial
improvement
• Non-responders
These profound insights deserve to be taken seriously and the van den
Bussche criteria may prepare the ground for future standards regarding not
only dementia treatment, but medical interventions in general.
H van den Bussche (2005) Problems and questions of pharmacological
therapy of Alzheimer’s disease with cholinesterase inhibitors. Z Gerontol
Geriat 38; Suppl 1: 18-20
Competing interests:
Fees from Eisai, Novartis,
Janssen-Cilag, Pfizer and others
Competing interests: No competing interests
„Generals“ commanding committees in the British (NICE) and German (I-
QUIP)
health systems are feared to triumph and withhold antidotes from those
suffering in the trenches, while irresponsible pseudo-science is the fuel
for
their perverse Zeitgeist (1). It is exactly this kind of desperate
statement from
an honourable colleague, which adds to our pain and shame, after the
publications of the British AD2000 study and a recent German exercise in
therapeutic nihilism (2). Potential malice and malingering of experts and
opinion leaders have been exposed by a fresh look from a group, unburdened
by previous experience. Their work was an innocent and absolutely
independent by-product of the “German Dementia Competence Network”,
sponsored by the German “Federal Ministry of Research”. A large number of
comments have been made, indicating that several important points were
missed. Here are a few more:
• published evidence clearly supports the symptomatic efficacy of
cholinesterase-inhibitors in (Alzheimer) dementia, but unpublished studies
with non-significant results may represent a real problem (they are
unavailable);
• results gathered from one population (e.g. from the USA) may not be
valid
for other types of populations;
• it is an illusion to assume that positive treatment effects can be
observed
reliably by an individual physician overseeing single patients or small
patient
groups, because of the heterogeneity of (Alzheimer) dementia and its
variable
course; this is one reason for carrying out large-scale, randomized,
double-
blind, controlled trials;
• the costs for effective anti-dementia substances are currently
high, - but
this must and will not be eternally so;
• we all want much more effective treatments for patients who already
are or
who will become demented in the near future, but these interventions may
not become available during our lifetimes (even though generals and their
propaganda of silver bullets have led the public to hold out without
proper
support for the last 15 years).
The authors (2) and Perry (3) stated that they were unaware of means
to
identify individuals who will respond to anti-dementia drugs. According to
the available evidence, patients with predominant cholinergic deficits
will
benefit most strongly from cholinesterase-inhibitors. Their clinical
features
are cognitive impairment plus:
• recurrent confusional states (“fluctuating course”) with
• visual hallucinations,
• marked EEG-slowing and
• particularly slim cholinergic nuclei in their high-resolution brain
-scans.
Some of these patients are said to suffer from the highly prevalent
“dementia
with Lewy-bodies” or a dementia with confusional states etc. Other
beneficiaries are students and airline pilots (PubMed knows more).
(1) D.G. Wilkinson. Evidence for cholinesterase-inhibitors in AD. BMJ
(2005)
(2) H. Kaduskiewitz et al. Cholinesterase-inhibitors for patients with AD.
BMJ
(2005)
(3) T.L. Perry. What is the clinically meaningful TNT/NHS. BMJ (2005)
Competing interests:
I received fees from Eisai,
Janssen-Cilag, Novartis, Pfizer,
and many others
Competing interests: No competing interests
A recent review published in the BMJ(1) questioned the conclusions of
the Cochrane evidence based reviews focussing on each of the 3 licensed
cholinesterase inhibitors that these agents conferred significant clinical
benefit.2,3,4 A subsequent letter presented qualitative data from a small
group of 22 carers, which consistent with the Kaduszkiewicz et al review,
suggested that the perceived benefits of treatment were modest.5 Whilst we
fully agree that it is imperative that systematic evidence from the
experiences of people with AD is available in the public domain and
contributes to the overall evidence base when appraising pharmaceutical
treatments; this information needs to be based on large systematic surveys
and not small case series. To ascertain the experiences of people with
dementia and their carers regarding treatment with anti-dementia drugs we
undertook a postal survey of people with dementia and their carers.
A questionnaire was posted to the membership of the UK Alzheimer's
Society and to memory clinics in the UK for distribution to people with
dementia and their carers, asking people with experience of the anti-
dementia drugs to respond. In total, 2,295 people who had experience of
one of the 3 licensed anticholinesterase drug treatments responded (2,060
- 77% donepezil, 474 (18%) rivastigmine, 487 (18%) galantamine).
There were 18 questions in the questionnaire, with topics including
the perceived benefits and side effects of drug treatment, access to and
information about the drug treatments and other care services used (full
questionnaire available at www.alzheimers.org.uk). A specific question
asked respondents “whether, taking everything into consideration, they
felt that the drug treatment they had received had worked” and to avoid
asking leading questions respondents were asked, in their own words, to
list up to 5 five ways in which treatment had been helpful.
Overall 1,569 (68%) of people who had been treated with at least one
of the anticholinesterase drugs stated that the treatment had worked. The
10 most frequently reported benefits included being more aware and more
active, calmer, taking more interest in things, improved conversation,
better quality of life and increased confidence, in addition to the
improvements usually evaluated in clinical trials.
These results are consistent with the findings of an Australian
consumer survey which found 70% of those who had experience of drug
treatments for dementia felt they were effective.6 Importantly the
benefits experienced extended well beyond stabilization of illness and
improvements in function and improved cognition and clearly illustrates
that the majority of people receiving cholinesterase treatment perceive
the therapies to be effective.
1) Kaduszkiewicz H, Zimmermann T, Beck-Bornholdt H, and van den
Bussche H. Cholinesterase inhibitors for patients with Alzheimer's
disease: systematic review of randomised clinical trials. BMJ 2005;331:
321 - 327
2)Olin J, Schneider L. Galantamine for dementia due to Alzheimer's disease
(Cochrane Review), In: The Cochrane Library, Issue 1, 2004.
3)Birks J, Grimley Evans J, Iakovidou V et al. Rivastigmine for
Alzheimer's disease (Cochrane Review), In: The Cochrane Library, Issue 1,
2004.
4)Birks JS, Harvey R. Donepezil for dementia due to Alzheimer's disease
(Cochrane Review), In: The Cochrane Library, Issue 1, 2004.
5)De Cauwer H G, August 11, 2005 response to: Kaduszkiewicz H, Zimmermann
T, Beck-Bornholdt H, and van den Bussche H. Cholinesterase inhibitors for
patients with Alzheimer's disease: systematic review of randomised
clinical trials. BMJ 2005;331: 321 – 327
http://bmj.bmjjournals.com/cgi/eletters/331/7512/321
6) Alzheimer’s Australia. Consumer medication survey. Alzheimer’s
Australia, 2005.
Competing interests:
CB has received consultancy fees from Pfizer, Novartis, Janssen and Shire, honoraria for presentations and research funding from Novartis and Janssen. JC, CB and SS are employed by the Alzheimer’s Society who actively campaign for fair access to effective drug treatments for people with dementia.AB has received expenses and honoraria from companies involved in the manufacture and marketing of drugs for Alzheimer’s disease. RH has no competing interests to declare other than being a psychiatrist who prescribes cholinesterase inhibitors for patients.
Competing interests: No competing interests
Sir,
Kaduszkiewicz et al (1) conducted a systematic review of clinical trials
of cholinesterase inhibitor treatments for Alzheimer's disease. They found
that these drugs conveyed benefits that were small in magnitude, and that
the clinical trials were anyway flawed in many regards. They questioned
the value of these drugs for Alzheimer's patients.
The review addressed only one part of the picture. Poor cognitive
benefits of the treatments notwithstanding, the cholinesterase inhibitors
are associated with substantial caregiver benefits. For example, in a
secondary analysis, Blesa (2) observed that, after 6 months of treatment,
the time spent by caregivers in assisting Alzheimer's patients was 1-1.5
h/day less with galantamine than with placebo, and the time spent in
supervising the patients was a further 1.5 h/day less. Thus, galantamine
saved the caregivers a total of 2.5-3 h/day.
In another secondary analysis of data from two 6-month,
randomized, double-blind, placebo-controlled trials conducted in 825
patients with Alzheimer's disease, Sano et al (3) showed that, relative to
placebo, galantamine saved caregivers an average of 32 min/day in terms of
assisting the patient with activities of daily living. The saving was 53
min/day in patients with moderate Alzheimer's disease. Furthermore,
relative to placebo, galantamine-treated patients could be left
unsupervised for an average of 27 min/day longer; this figure was 68
min/day in patients with moderate Alzheimer's disease.
In this context, it should be noted that galantamine-related
caregiver benefits are related to the neuropsychiatric benefits of the
drug, and not to its cognitive benefits (4).
References
1. Kaduszkiewicz H, Zimmermann T, Beck-Bornholdt H-P, van den Bussche H.
Cholinesterase inhibitors for patients with
Alzheimer's disease: systematic review of randomised clinical trials. BMJ
2005; 331: 321-327.
2. Blesa R. Galantamine: therapeutic effects beyond cognition. Dement
Geriatr Cogn Disord 2000; 11 (suppl 1): 28-34.
3. Sano M, Wilcock GK, van Baelen B, Kavanagh S. The effects of
galantamine treatment on caregiver time in Alzheimer's disease. Int J
Geriatr Psychiatry 2003; 18: 942-950.
4. Cummings JL, Schneider L, Tariot PN, Kershaw PR, Yuan W.
Reduction of behavioral disturbances and caregiver distress by galantamine
in patients with Alzheimer's disease. Am J Psychiatry 2004; 161: 532-538.
Competing interests:
I have been a site investigator in multicenter clinical trials.
Competing interests: No competing interests
Many randomized controlled trials of the three cholinesterase
inhibitors, donepezil, galantamine and rivastigmine, have been completed
and published. Cochrane reviews of the three drugs, an individual patient
data meta-analysis of donepezil by a joint team from the manufacturing
company and independent scientists, and a meta-analysis of all three have
been published. The combined evidence concludes that, for each of the
three drugs, there are benefits for patients with Alzheimer’s disease. The
three drugs are licensed for use in many countries including those of
Europe, and North America. In the UK NICE has recommended their use, based
on the results of trials and systematic reviews, although this is now
under review on account not of efficacy but of cost effectiveness.
Kaduszkiewicz et al dismiss all this evidence as worthless, because
they state that without exception the trial methodology is poor and all
the previous reviews are not evidence based. Are they just being
vexatious? The reviews criticised included a systematic search for trials,
assessment of the quality of the evidence and a synthesis of the data
using meta-analysis. What is their definition of evidence based?
In their own systematic review of these drugs, discussion with
reference to a checklist of points led to qualitative assessment of the
methodology of the trials. They criticise the Cochrane reviews for failing
to assess quality and thence to recognize the methodological shortcomings
which they have detected. They make the following points and mostly they
are misguided.
1. No correction for multiple outcomes.
This could be considered a valid criticism. But we are not seeing a few
random significant treatment effects, nearly all trials show a significant
effect for the drug for the primary outcomes, and these are highly
significant in the meta-analyses of the Cochrane reviews. No correction
for multiple comparisons would change this. Kaduszkiewicz et al do not
attempt a synthesis of data across the trials.
2. Missing intention to treat analyses and incomplete data.
The loss of patients from the trials is a problem. In the Cochrane meta-
analyses of donepezil (Birks and Harvey 2005), we do not find a
significant difference between treatment and placebo for the number of
patients who left the trial (dropouts), but a difference is found for
rivastigmine (Birks et al 2005) and higher dose galantamine (Olin and
Schneider 2005). In order to investigate the effect dropouts may have on
the results, the Cochrane reviews report and discuss the analyses of the
intention to treat (ITT), the completers (OC) and where available (e.g.
the rivastigmine review) the OC plus retrieved dropouts (OC+RDO)
populations. By comparing the analyses it is possible to interpret some of
the effects of dropouts. This is the best that can be done when one has no
access to individual patients’ data.
It is inaccurate to state that dropout rates were not considered and did
not influence the conclusions in the Cochrane review of rivastigmine.
3. Different design and methodological flaws.
There were differences in the study designs, often because they were
addressing different questions, not because of flawed design. For example
Kaduszkiewicz et al criticise Mohs (2001), the primary endpoint of which
was time to clinically evident decline in function, because patients left
the trial at endpoint, although the design met the objective of the trial.
A Cochrane meta-analysis only pools results when the studies are
considered comparable. The Cochrane review discusses the designs of all
trials and, incidentally, does not include Mohs (2001) in the main meta-
analyses.
4. Imbalance at baseline.
If the differences between groups for many characteristics are examined at
baseline there will almost certainly be a few statistically significant
differences; it is usually a pointless exercise. It is not evidence of
poor methodology. The authors themselves know the problems associated
with testing for multiple outcomes, but they forget to take their own
advice on multiple testing when they state that there was an imbalance at
baseline (eg. height in Rogers 1996).
Having dismissed the trials as too poor for consideration, and the
apparent treatment effects as too small to be of any use, they conclude
that the benefits are not proven. But then, in an about turn, they also
conclude that adverse effects are without doubt present. Their criticism
of the methodology of the trials and of inferences due to testing multiple
outcomes without correction are all forgotten. It is to avoid decisions
resting on such inconsistency of approach that the Cochrane Collaboration
produces high quality systematic reviews.
Jacqueline Birks
Co-ordinating Editor
Cochrane Dementia and Cognitive Improvement Group,
Division of Geratology,
Department of Clinical Medicine,
University of Oxford,
Radcliffe Infirmary,
Oxford
Mohs RC, Doody RS, Morris JC, Ieni JR, Rogers SL, Perdomo CA, Pratt
RD. A 1-year, placebo-controlled preservation of function survival study
of donepezil in AD patients. Neurology 2001;57(3):481-8.
Rogers SL, Friedhoff LT and the Donepezil Study Group. The Efficacy and
safety of Donepezil in patients with Alzheimer's disease: results of a US
multicentre, randomised, double-blind, placebo-controlled trial. Dementia
1996;7:293-303.
Birks J, Harvey R The efficacy of donepezil for mild and moderate
Alzheimer’s disease (Cochrane Review). In: The Cochrane Library, issue 2,
2005. Chichester, UK: John Wiley
Birks J, Grimley Evans J, Iakividou V, Tsolaki M . Rivastigmine for
Alzheimer’s disease (Cochrane Review). In: The Cochrane Library, issue 2,
2005. Chichester, UK: John Wiley
Olin J, Schneider L .Galantamine for Alzheimer’s disease (Cochrane
Review). In: The Cochrane Library, issue 2, 2005. Chichester, UK: John
Wiley
Competing interests:
None declared
Competing interests: No competing interests
Kaduszkiewicz and her colleagues are right to realise that the
conclusions which they draw about cholinesterase inhibitors (1) are out of
step with those of other reviewers. Many of the imperfections that they
so astutely identified are omissions in the reporting of the trials in
journals, not defects in quality of the trial. Their response, which is
to write off the results of those trials as non-informative, is not
justified and certainly not made from the perspective of evidence-based
medicine.
Cochrane reviewers have not considered lack of good reporting in this
area – as frustrating as that is – to be a reason to exclude trials
results from further consideration. Rather they have sought additional
material such as protocols and clinical study reports from the drug
companies and have asked them for more information. Moreover, most, if
not all the trials that Kaduszkewicz questions were carried out under Good
Clinical Practice and International Conference of Harmonization guidelines
(e.g. http://www.fda.gov/oc/gcp/guidance.html) in order to meet the
standards necessary for regulatory review. These trials in fact meet the
quality standards that Kaduszkiewicz and colleagues sought but did not
find in the published reports.
Their criticism of the failure of analyses to correct for multiple
testing is misplaced and rests on their misunderstanding of the trials’
methodology. The oft cited Bonferroni correction is the most conservative
of several post hoc adjustments for multiple comparisons that can be used
in the absence of pre-specified primary outcomes. However, all the trials
they included in their review had clearly specified primary outcomes,
usually two co-primaries, a cognitive outcome and a global, each having to
reach a p < 0.05 alpha error level. Notwithstanding this error, their
paper fails to refer to any meta-analytic results. Examination of
Cochrane meta-analyses shows that the effects of cholinesterase inhibitors
would remain statistically significant even if Bonferroni corrections were
applied to the multiple pooled outcomes in the meta-analyses.
Their point about differential dropout is already widely rehearsed
and understood. They are wrong to suggest that this was inadequately
considered in the conclusions of Cochrane reviews. These reviews show
that differential dropout occurs for rivastigmine and higher dose
galantamine but not donepezil. It was carefully interpreted by comparing
the last observation carried forward and completer analyses, and by
retrieved dropouts analyses, when part of the protocol, so as to give the
best approximation to true intention to treat analyses where available.
This is the best that can be achieved without access to individual patient
data. Altogether this constitutes evidence that needs to be integrated
into clinical assessments of effectiveness and utility, and not dismissed
out of hand.
The criticism of the Mohs et al trial(2) – that investigators could
remove a patient on the basis of the investigator’s clinical judgement
which can be subjective - is tautological and gratuitous. The judgement
was to be applied in interpreting definitions of ‘clinically evident
decline’. Clinical judgement is what physicians do. This trial met its
own objectives – assessment of time to loss of daily activities - which
were not the same as most of the other trials.
Most of these trials were designed on to show differences in mean
cognitive and global function between drug and placebo groups and were not
designed or powered to test post hoc for subgroups or prospectively
designed to define individual responders. The authors’ comments about
subgroups of responders and the need to define assessment procedures which
distinguish responders are sensible, but they are not new and bear no
relation to the text in the preceding sections.
There is no such thing as a perfectly reported trial. But to write
off the available trials as being of ‘poor’ methodological quality for the
reasons offered does not adequately deal with the evidence and does
nothing to promote the standing of systematic reviews - of which this was
not a good example.
1. Kaduszkiewicz H et al. Cholinesterase inhibitors for patients
with Alzheimer's disease: systematic review of randomised clinical trials
BMJ 2005; 331: 321-327
2. Mohs RC et al A 1-year, placebo-controlled preservation of function
survival study of donepezil in AD patients. Neurology. 2001 Aug
14;57(3):481-8.
Competing interests:
RM is paid to coordinate production of reviews by the Cochrane Dementia and Cognitive Improvement Group, has received speaking fees, support for attending conferences and support for a course he runs from all manufacturers of cholinesterase inhibitors. LS has received speaking and consulting fees from all manufacturers of cholinesterase inhibitors.
Competing interests: No competing interests
According to Kaduszkiewicz et al. the scientific basis for
recommendations of acetylcholinesterase (AChE) inhibitors for the
treatment of Alzheimer’s disease is questionable. However we are
concerned that a significant proportion of the population suffering from
Alzheimer’s dementia appears to have been forgotten – those with learning
disabilities.
It is well documented that 54.5% of people with Down’s Syndrome aged
60-69 years are affected by Alzheimer’s dementia, compared to only 5% of
the general population aged over 65 years.
The original NICE guidelines (2001) approved the use of AChE
inhibitors for the management of Alzheimer’s dementia in the context of
specialist assessments and monitoring. The mini mental state examination
(MMSE) was recommended as the most appropriate tool to determine treatment
initiation and discontinuation. Although NICE did not specifically
exclude people with learning disabilities, the MMSE is outside the normal
range in this population.
This has meant that the use of AChE inhibitors in people with Down’s
Syndrome and dementia has been limited compared to the general population.
In addition, due to important and appropriate issues regarding giving
informed consent, people with learning disabilities are usually excluded
from clinical trials. (As far as we are aware there are no published
multicentre, double blind randomised controlled trials on the use of AChE
inhibitors in people with Down’s syndrome.) Finally, it is well
recognised that people with learning disabilities metabolise and respond
to drugs differently (and not always predictably) from the general
population. Therefore guidelines derived from double blind, randomised
controlled trials should not solely be used to make decisions about the
use of the drugs in people with learning disabilities.
Our local SHA agreed to fund the use of AChE inhibitors in people
with Down’s Syndrome and Alzheimer’s dementia provided a suitable audit
tool was developed to determine the effectiveness of treatment. A Trust
Down’s and Dementia care pathway was therefore developed. People with
Down’s syndrome over the age of 35 years and known to the service are
screened every five years using the Assessment of Motor Processing Skills
(AMPS) and Dementia Questionnaire for persons with Mental Retardation
(DMR). These were chosen as the most appropriate tools due to their
sensitivity to change. If the client develops dementia these assessments
are carried out every 3-6 months for monitoring of disease progression and
treatment effectiveness.
A recent audit of the pathway revealed everybody with Alzheimer’s
dementia (n=30) was considered for treatment and 24 (80%) were commenced
on an AChE inhibitor. The reasons for not commencing treatment were that
the client’s dementia or their premorbid level of functioning was so
severe that treatment would not produce worthwhile benefits and in one
case only, because of concerns regarding gastrointestinal side effects.
Of those who had commenced treatment 22 (92%) remained on treatment. It
was continuing to be effective for 12 (50%) of clients (time from
initiation 1-45 months), was too early (i.e. < 3 months since
commencing treatment) to decide on effectiveness for 6 (27%) of clients
and the effectiveness was being questioned in 5 (23%) of clients (time
span 4 -22 months). For the two people who had stopped treatment, one had
died of causes unrelated to the Alzheimer’s dementia and the other had
gastrointestinal side effects (donepezil). Only one other client had
suffered side effects – diarrhoea on commencing donepezil and this had
abated when the dose was reduced and increased at a slower rate. The
usual daily starting doses were 5mg for donepezil, 6mg for rivastigmine
and 8mg for galantamine.
In conclusion, since NICE (2001) did not specifically consider the
needs of people with learning disabilities those of us working in the
speciality have had to devise more appropriate ways of monitoring the
effectiveness of AChE inhibitors in order not to exclude and therefore
discriminate against people with learning disabilities having access to
this treatment. Using clear assessments there is a significant
proportion of people within our audit who have benefited from these drugs.
We are concerned that NICE will fail to realise the need to consider
other levels of evidence for people with learning disabilities. Therefore
if NICE do recommend not using AChE inhibitors, people with learning
disabilities could again be discriminated against.
Competing interests:
None declared
Competing interests: No competing interests
BMJ published an important paper by Kaduszkiewicz, Zimmermann, Beck-
Bornholdt and van den Bussche in their 06 August edition. (1) The paper,
“Cholinesterase inhibitors for patients with Alzheimer's disease:
Systematic review of randomised clinical trials”, is a meta-analysis of 22
trials. It demonstrated that flawed methodology and negligible clinical
benefits undermine previously asserted recommendations for some of the
commonly prescribed anticholinesterase inhibitors for Alzheimer’s disease.
Separately, an important 2002 study by Ann Hamer and her associates
published in the Journal of Managed Care Pharmacy (JMCP) revealed similar
findings of flawed methodology and limited clinical benefits. The subject
of that investigation was the “off-label” use of gabapentin, a drug
approved for adjunctive use in partial seizure disorders and post-herpetic
neuralgia but not for the wider variety of chronic pain and psychiatric
syndromes which accounted for much of its off-label prescribing.
In that study, the authors retrospectively reviewed medical records
from 105 patients who were treated with gabapentin, prescriptions
reimbursed by Oregon Medicaid. (2) 95% of the gabapentin prescriptions
were off-label. More significantly, 88% of patients receiving gabapentin
did not demonstrate a positive response.
Unfortunately for those who paid the bills for gabapentin therapy and
for those consumers who reported no benefit and/or unpleasant side-
effects, these published negative results came almost 9 years after the
successful release and marketing of gabapentin. [Gabapentin (Neurontin-
Parke-Davis) was approved by the US FDA on 30 December 1993. In 2001,
according to Hamer et al, gabapentin was in the top-10 drug products by
total costs paid by Oregon Medicaid fee-for-service. ]
One of the likely sources of over-enthusiastic off-label prescribing
and lack of documented off-label benefits has been the uncritical adoption
of claims of benefits based upon open-label trials or inadequately
populated cohorts. Another occurs when research scientists are not
expected to compare their original findings with those of longer term
studies which contradict their original conclusions.
Some of these problems were recently described in an investigative
report by John Ioannidis that appeared in the 13 July 2005 edition of the
Journal of the American Medical Association (JAMA.) (3) He reviewed 49
highly cited original clinical research studies to determine whether
subsequent analysis could reiterate or replicate original positive
findings. Although some studies (24%) went unchallenged, about 32% of
studies----or nearly one-third of preliminary highly touted research
findings involving various medicines and therapeutics---did not have the
reliability and validity of their findings confirmed by follow-up
research, with some outcomes specifically contradicted by subsequent
studies.
Kudos to BMJ, JAMA, and JMCP (and other comparably reputable peer-
reviewed journals) for identifying research whose original findings have
subsequently been seriously qualified or invalidated. We believe ethical
journals have an obligation to expose research making claims of efficacy--
–especially for off-label use of medications---when these claims do not
withstand subsequent prospective scrutiny. Pharmaceutical journals have a
particularly strong obligation to do this since they hold a special trust
in the provider community, often being the first place where new drugs are
introduced and off-label use is promoted. Furthermore, we maintain that,
in order to repair public and provider trust in academic and
pharmaceutically-sponsored research, journals must expand their exposure
of data analyses that are later found to have been biased and/or
predicated upon misstatements regarding the reliability and validity of
prior published results.
Even more carefully considered research and publishing criteria still
fall short of protecting the patient and provider communities when
publication is too-long delayed. Timing is keenly important.
Indeed, even after the publication of spurious data collection or
misrepresented results, multi-year lapses routinely occur between non-
confirmatory re-analysis and bogus preliminary findings. Therefore, if
journal and media scrutiny are effectively to minimize the ramifications
(morbidity, mortality and added healthcare costs) of questionable data,
published re-analysis must occur, if at all possible, in close temporal
relationship to the original publications.
Stefan P. Kruszewski, M.D.
Jeffrey A. Brown, M.D., J.D., M.P.H.
Clinical Associate Professor of Psychiatry
University of Medicine and Dentistry of New Jersey
1 Kaduszkiewicz H, Zimmermann T, Beck-Bornholdt HP, van den Bussche
H. Cholinesterase inhibitors for patients with Alzheimer's disease:
systematic review of randomised clinical trials BMJ. 2005 Aug 6;
331(7512): 321-327
2 Hamer, AN M, Haxby, DG, McFarland, BH and Ketchum, K. Gabapentin
Use in a Managed Medicaid Population. JMCP 2002; 8(4): 266-271
3 Ioannidis, JP. Contradicted and Initially Stronger Effects in
Highly Cited Clinical Research. JAMA. 2005; 294 (2): 218-228
Competing interests:
None declared
Competing interests: No competing interests
Re: Cholinesterase inhibitors for patients with Alzheimer's disease: systematic review of randomised clinical trials
Dear Editor,
Please consider recent meta-analysis of Cholinesterase inhibitors and Memantine, suggesting that cognition (as assessed by ADAS-Cog) deteriorated quicker on these drugs (mostly both) compared to people not given these drugs. fits in with the original (most naturalistic) AD 2000 and the initial NICE technology appraisal.
In addition, we need to get rid of the dogma of 'living well with dementia'; is a joke, a cruel one at that, as most carers will tell you. Thinking in terms of dementia being a life limiting illness needing a palliative approach a better communication strategy.
Reference.
Association of Concomitant Use of Cholinesterase Inhibitors or Memantine With Cognitive Decline in Alzheimer Clinical Trials: A Meta-analysis. Richard E. Kennedy, MD, PhD1; Gary R. Cutter, PhD1; Mackenzie E. Fowler, MPH1; et al Lon S. Schneider, MD, MS2
JAMA Netw Open. 2018;1(7):e184080. doi:10.1001/jamanetworkopen.2018.4080
Competing interests: No competing interests