Alzheimer’s & Dementia 2 (2006) 118 –125

Review Article

Do cholinergic therapies have disease-modifying effects in

Alzheimer’s disease?
Marwan N. Sabbagha,*, Martin R. Farlowb, Normal Relkinc, Thomas G. Beachd
a
Cleo Roberts Center for Clinical Research, Sun Health Research Institute, Sun City, AZ, USA
b
Department of Neurology, Indiana University School of Medicine, Indianapolis, IN, USA
c
Department of Neurology, Weill Cornell Medical School, New York, NY, USA
d
Civin Laboratory for Neuropathology, Sun Health Research Institute, Sun City, AZ, USA

Abstract The most widely studied and used therapies for Alzheimer’s disease (AD) are based on improving
cholinergic function in the central nervous system. The acetylcholine-esterase inhibitors (ChEIs) tacrine,
donepezil, rivastigmine, and galantamine are all approved, and the latter three are widely used for the
symptomatic treatment of mild to moderate AD. Recent research has found that these drugs may act by
a variety of other mechanisms including inhibition of butylcholinesterase, regulation of nicotinic recep-
tors, decreasing amyloid precursor protein (APP) and A␤ production, and regulation of tau phosphor-
ylation that may influence disease progression. There is also emerging evidence from clinical trials that
the ChEIs may delay cognitive and functional progression. Other cholinergic drugs such as muscarinic
agonists have been explored, and although they are not approved, there is robust preclinical evidence for
a beneficial, perhaps disease-modifying effect. This review summarizes evidence suggesting that these
drugs may do more than improve symptoms; they may delay biological progression of the disease.
© 2006 The Alzheimer’s Association. All rights reserved.
Keywords: Muscarinic; Cholinesterase inhibitors; Alzheimer’s disease; Prevention; Therapy

1. Introduction disease change over time in AD. The first model is symp-
tomatic in which a medication is administered to a patient,
Cholinergic drugs are the most commonly prescribed drugs and their symptoms improve for a period of time, but the
for Alzheimer’s disease (AD). There is an extensive body of rate of decline is not affected, and ultimately the downward
literature focused on the efficacy of cholinergic therapies and trajectory is parallel to the untreated condition.
their utility in clinical practice. This debate has far-reaching In contrast, disease-modifying medications are intended
implications because medical decision making and pharmaceu- to affect the trajectory of decline such that over time the
tical benefits are based on the current perceptions posited in the treated AD patient continuously shows a lesser rate of
literature. This review is intended to explore the preclinical and decline and a progressively greater preservation of cogni-
clinical evidence that cholinergic drugs may have disease- tion and function compared with the untreated patient. Stan-
modifying effects and that perhaps the drugs may have pro- dard instruments performed serially that measure change in
longed and sustained effects over time. cognition such as the Mini Mental State Examination
(MMSE), the Dementia Rating Scale (DRS), or the Alzhei-
2. Symptomatic Effects versus Disease-Modifying mer’s Disease Assessment Scale (ADAS-cog) should be
Effects able to measure this.
The current belief is that cholinomimetic medications
improve symptoms only. There are two models for tracking
3. Cholinergic Changes Occur Early in AD

*Corresponding author. Tel.:623-875-6500; Fax: 623-875-6504. Cholinergic therapy may be the most physiologically
E-mail address: [email protected] appropriate therapy for AD because cortical cholinergic
1552-5260/06/$ – see front matter © 2006 The Alzheimer’s Association. All rights reserved.
doi:10.1016/j.jalz.2006.02.001
 M.N. Sabbagh et al. / Alzheimer’s & Dementia 2 (2006) 118 –125 119

deafferentation is a known accompaniment of normal aging neurotransmitter networks that ordinarily would deteriorate
[1] and is probably an early, preclinical event in AD [2–5]. with loss of cholinergic innervation, and effects on cerebral
The cortical cholinergic deficit begins in midlife [1] and is blood flow [20]. We will limit our discussion to the effects
closely followed by an increase in cortical A␤ levels [6]. of cholinergic agents on A␤, because this is the hypothesis
Two-thirds of all humans have cortical A␤ deposits by age with the most preclinical data supporting a plausible mech-
70 [7]; by age 90, this increases to more than 90%. Cortical anism.
A␤ deposits in nondemented aging humans are associated Activation of muscarinic cholinergic receptors may reg-
with a more pronounced cholinergic deficit, relative to those ulate ␤APP metabolism. In vitro studies have found that
lacking or having few 〈␤ deposits [2–5]. Acetycholines- muscarinic M1 and M3 agonists stimulate processing of
terase but not butylcholinesterase has been shown to be a ␤APP via the nonamyloidogenic pathway [21,22]. This re-
potent amyloid-promoting factor [8]. Others have suggested sults in increased release of sAPP␣ or reduced A␤ produc-
that the cholinergic deficit occurs later in AD, at the stage of tion [21–33]. Agents used include carbachol, a nonselective
moderate or even severe dementia [9,10]. Some have found muscarinic agonist [22,27,32,34]; AF102B, a selective M1
that the nucleus basalis of Meynert (NBM) neurons con- ligand [26,27]; xanomeline, which is an agonist at both the
taining ChAT and vesicular acetylcholine transporter were M1 and M4 sites [31]; and Lu 25-109, which is a combined
relatively preserved in mild cognitive impairment (MCI) M1 agonist and M2 antagonist, previously in drug develop-
and early AD patients. However, pretangle, AT8, and ment for the treatment of mild to moderate AD [30,35].
Alz-50 cytopathology were present in NBM neurons of Application of these M1 or M3 muscarinic agonists resulted
MCI subjects. In addition, there was a loss of pan-neuro- in increased secretion of sAPP␣ or decreased secretion or
trophin receptor p75ntr, and nerve growth factor–specific production of A␤. Selective M1 agonists were reported in
receptor tyrosine kinase immunoreactivity, which are ex- one study as being more potent at stimulating sAPP␣ se-
pressed by the NBM, and are anterogradely transported to cretion than nonselective muscarinic agonists [31]. In con-
bind the nerve growth factor. These studies are difficult to trast, experiments using cells expressing only M2 and M4
reconcile. Some of them appear to indicate a preservation of receptors found that activation of these receptor subtypes
the cholinergic system in MCI and early AD patients, had no effects on sAPP␣ secretion [22]. At least three
whereas others found major structural damage to the NBM. studies, however, have confirmed that secreted A␤ levels
These studies are disparate partially because clinical, rather are decreased [23,29,31].
than histopathologic staging of AD was used, resulting in The selectivity for the M1 and M3 receptors suggested
the mixing of subjects with and without significant AD that the effects on ␤APP processing are mediated via acti-
histopathology. vation of phospholipase C, formation of diacylglycerol, and
Cholinergic agents may be uniquely suited to AD for activation of protein kinase C (PKC) [25]. Mitogen-acti-
disease modification because animal experiments have sug- vated protein kinase-dependent (MAPK) [33] and Wnt path-
gested that cortical cholinergic deafferentation may be the ways [19] have also been proposed to be involved. PKC
cause of age-related A␤ deposition [4]. Rats with lesions of inhibitors block the muscarinic agonist-induced secretion of
the nucleus basalis magnocellularis (nbm), which supplies sAPP␣ [22], whereas inducers of the enzyme cause in-
the cerebral cortex with its cholinergic innervation, have creased secretion [21,23,36]. Induction of PKC decreases
increased cortical concentrations of ␤APP message and secretion or production of A␤ [23,36 –38]. The final link in
protein [11,12] and decreased cortical [13] and cerebrospi- the degradative pathway is most probably mediated by in-
nal fluid (CSF) [14] levels of sAPP␣. Rabbits with lesions hibition or stimulation of one or more of the three enzymatic
of the nbm have 2.5-fold and 8-fold elevations of cortical activities (␣, ␤, ␥) involved in the cleavage of ␤APP,
〈␤ 40 and 42, respectively, and develop vascular and although this has not yet been proven. Additionally, the
perivascular A␤ deposits [4,15,16]. Control animals with ChEI phenserine has been reported to reduce cell secretion
lesions of the cortically projecting noradrenergic locus cer- of soluble ␤APP and A␤ by blocking the promoter form of
uleus do not accumulate A␤. ␤APP, which results in down regulation of its mRNA tran-
script [39,40]. In cell culture, tacrine decreased soluble
4. Preclinical Evidence Supporting a Disease ␤APP levels in conditioned media, which was not affected
by cotreatment with muscarinic agonists, suggesting that
Modifying Effect for Cholinergic Agents
this effect does not depend on cholinesterase activity [41].
Several plausible explanations for a disease-modifying The extent to which cholinergic-modulated APP process-
effect of cholinergic agents, including ChEIs, have been ing is operative in the brain has been examined. Carbachol
proposed. The first mechanism proposed includes reduction was found to cause increased release of sAPP␣ but only in
of cortical A␤ concentrations, an effect that has been shown the presence of gallamine, an M2 antagonist, suggesting that
by multiple in vitro and animal model studies [15,17,18]. M2 stimulation inhibits sAPP␣ secretion, a premise that
Other possible mechanisms include protection against tau received additional support from the observation that gal-
aggregation [19], delay of secondary dysfunction in other lamine administration resulted in increased release [42].
120 M.N. Sabbagh et al. / Alzheimer’s & Dementia 2 (2006) 118 –125

The muscarinic agonist WAL 2014 stimulated sAPP␣ re- data suggest that ChEIs can influence APP trafficking in
lease only at lower doses, which was consistent with its peripheral cells and in cell culture.
biphasic nature, being selective for M1/M3 receptors at Other animal studies have shown that cholinergic ther-
lower doses but also stimulating M2 and M4 receptors at apy prevents experimentally induced A␤ deposition. Treat-
higher concentrations [42]. In other studies [43,44], differ- ment of a rabbit model of vascular A␤ deposition with
ent ChEIs (physostigmine and 2,2 – dichloro-vinyl dimethyl either physostigmine or AF267B prevented the biochemical
phosphate) all caused increased secretion of sAPP␣, but and histologic accumulation of cortical A␤ [14,15,16],
metrifonate did not [45], whereas bethanechol, a nonspecific whereas nicotine [58,59] and AF267B have both prevented
muscarinic agonist, had the same result. Lu 25-109 [35], A␤ accumulation in transgenic AD mouse models.
carbachol, and AF102B [46] have all been shown to stim- ChEIs may have neuroprotective properties separate
ulate the ␣-secretory pathway. from muscarinic activity. Donepezil, rivastigmine and ga-
In vivo studies support in vitro experiments. The linkage lantamine have been shown in vitro to protect against A␤
between PKC activation and resultant sAPP␣ secretion has toxicity and okadaic acid toxicity in SH-SY5Y cells [60].
received support from an in vivo experiment in which phor- Galantamine may, in addition to blocking A␤1-40 toxicity,
bol ester was injected directly into mouse brains [36]. Treat- prevent apoptotic cell death by increasing expression of
ment of animals with lesions of the nucleus basalis magno- Bcl-2 [61]. Galantamine also blocks A␤-enhanced gluta-
cellularis with RS-86, a muscarinic agonist with partial mate toxicity in cultured rat neurons [62].
selectivity for the M1 receptor, increased CSF levels of The studies summarized above support the concept that
sAPP␣ [14]. Phenserine has an effect contrary to that pre- cholinergic agents may have anti-amyloidogenic actions,
dicted by in vitro work by decreasing CSF levels of sAPP mediated by several potential molecular mechanisms. Be-
[47]. cause accumulation of A␤ is hypothesized to be the com-
In vivo studies [15,48,49] have directly measured CNS mon cause of the various forms of AD, cholinergic therapy
A␤ concentrations after treatment of normal animals with has the potential to delay progression of the disease by
cholinergic agents. The M1-selective muscarinic agonists slowing or halting the accumulation of A␤. As stated be-
from the AF series: AF102B, AF150(S), and AF267B [49] fore, cholinergic agents may be uniquely suited to AD for
significantly reduced CSF levels of A␤ 40, and two of the disease modification, because animal experiments have sug-
three agents significantly reduced CSF and cortical levels of gested that cortical cholinergic deafferentation may be the
A␤1-42. Guinea pigs treated with the ChEI physostigmine cause of age-related A␤ deposition [4]. Rabbits with lesions
had significantly lowered cortical concentrations of insolu- of the nbm have 2.5-fold and eight-fold elevations of cor-
ble (formic acid-extracted) A␤1-40[48]. tical 〈␤ 40 and 42, respectively, and develop vascular and
Cholinergic agents have been shown to lower blood A␤ perivascular A␤ deposits [4,15,16]. Treatment of lesioned
concentrations in humans and human cultured cell lines. rabbits with AF267B or physostigmine decreases A␤ dep-
One group compared AF102B and physostigmine and found osition and CNS A␤ concentrations [15]. These results
that the M1 agonist lowered blood A␤, whereas the ChEI show that cholinergic inactivity leads to accumulation of
did not [50,51]. Another study found, in contrast, that blood A␤, whereas treatment with muscarinic agonists or ChEIs
concentrations of tacrine were correlated significantly with may partially reverse this effect.
decreased plasma A␤ [52]. However, this difference may be The experimental work summarized here suggests that
owing to tacrine’s additional reported action on the APP cholinergic agents might simultaneously provide symptom-
promotor. In cell culture of human neuroblastoma cells, atic benefits in AD while having actions that may compen-
tacrine administration reduced levels of A␤40 and A␤42 sate for the proamyloidogenic effects of progressive cholin-
[53]. In cell cultures of human neuroblastoma, treatment of ergic deficiency, which occurs with increasing severity
the cells with donepezil released significantly more soluble throughout the disease process.
APP␣ into the medium [54]. In a clinical trial of AD
subjects assigned randomly to donepezil or no treatment, 5. Clinical Evidence of Disease-Modifying Effects of
the ratio of platelet APP isoforms (130 kDa/ 106-110 kDa) ChEIs
was measured. After 30 days of treatment, subjects with
donepezil had significantly higher ratios [55]. The treatment There is growing clinical evidence to suggest that ChEIs
effect seems to be influenced by the apolipoprotein E ge- have disease-modifying effects and are not merely pallia-
notype with non-⑀4 carriers having the more robust change tive, as has long been assumed. A recent meta-analysis of
in APP isoform ratios compared with ⑀4 carriers [56]. A longer-term clinical trial data [63] suggests that long-term
similar study was done with galantamine. AD subjects slowing, but not cessation of, cognitive decline with ChEIs
treated with galantamine for up to 12 weeks showed a may occur.
significantly increased ratio of high molecular weight iso- The Stern model [64] predicts a rate of decline for AD
forms [130 kilodaltons (kDa)] to low molecular weight patients between seven and 11 points per year on the Alz-
isoforms [106 and 110 kDa] of platelet APP [57]. These heimer’s Disease Assessment Scale-cognitive subscale
 M.N. Sabbagh et al. / Alzheimer’s & Dementia 2 (2006) 118 –125 121

(ADAS-cog). This is larger than the decline predicted using ADAS-cog score at entry into the placebo-controlled studies
a baseline-dependent mathematical algorithm for “un- was 24.6. Patients remaining on rivastigmine treatment for
treated” ADAS-cog decline, and larger than the average up to five years crossed the threshold to moderately severe
annual decline reported for untreated patients in the litera- dementia at about 2.5 years and did not decline to severe
ture (approximately seven points) [65]. Thus, it is probably dementia. The mean ADAS-cog score was 36.8 (indicating
wisest to use all of the above estimates to ascertain whether “moderately severe” AD) in these patients at 260 weeks,
a disease-modifying effect occurs with long-term treatment and the average annual decline was approximately 3.9
with ChEIs. ADAS-cog points per year, a rate of decline considerably
Indications of changing trajectory came from initial 12- slower than has been seen in natural course studies or
month studies of donepezil where the treated group had an placebo groups of drug studies. A similar analysis done on
annualized rate of decline on the MMSE of 1.5 points per the MMSE (baseline score ⫽ 19.3) showed that AD subjects
year versus 3.7 points per year in the untreated group [66]. on rivastigmine for five years maintained an MMSE score
A randomized placebo-controlled international study of the of greater than 10, whereas the projected placebo decreased
effects of donepezil compared with a real placebo group for below a score of 10 points [74].
12 months showed a 2.3-point difference in the MMSE Long-term data from studies of galantamine also suggest
between treatment and placebo [67]. Similar effects were disease modification. Tariot and Kershaw [78] followed up
seen at 18 months [68]. Analysis of ADAS-cog data in the with AD subjects taking galantamine for 18 months. They
98-week open label extension of one of the US randomized show that galantamine users at the maximal dose declined
studies showed that the expected rate of decline was half only two points on the ADAS-cog over 18 months. Pirttila
that of the untreated group [69]. Doody et al [70] analyzed et al [79] reported 36-month open-label extension data of
open-label extension from two randomized US trials for 144 the subjects that were in the three- or six-month randomized
weeks. The ADAS-cog declined four points per year in the placebo-controlled trial of galantamine. After 36 months,
donepezil-treated group compared with six points in the long-term users of galantamine declined 4.1 points per year
placebo group, which suggests a decreased rate of disease on the ADAS-cog compared with the expected 7.3 points
progression in the treated group. In another open-label study per year with the projected placebo. In the US trials, the
of 133 subjects assessing effects up to 4.9 years, Rogers et 36-month data showed that the rate of decline was 3.4 points
al [71] found that the donepezil-treated group declined an per year on the ADAS-cog, approximately 50% less than the
average of 4.5 points per year on the ADAS-cog, less than projected rate of decline [80].
the expected decline in a placebo-only group. Interpretation of long-term and open-label data may be
Retrieved dropout data from rivastigmine trials suggest biased. Farlow et al [81] reported that only 25% of subjects
that subjects that received rivastigmine and discontinued the enrolled in the original rivastigmine ENA713 B352 were still
medication early had less cumulative decline on ADAS-cog taking rivastigmine after five years. Many studies reporting
versus either the placebo patients who continued in the long-term data of ChEIs suffer from a significant attrition rate
blinded study or the placebo dropouts. Leber and others of up to 90% of the original sample. This attrition may well be
[72,73] have suggested that continued beneficial effects in selective with patients who experience more rapid progression
treated patients after drug withdrawal suggests an effect on of their symptoms being more likely to stop taking their ChEI.
diffuse plaque. Small et al [74] followed up with AD sub- However, data must be interpreted with caution because they
jects taking rivastigmine for five years. From entry into the come mostly from open-label studies [63]. Nevertheless, there
placebo-controlled studies, patients remaining on rivastig- is a general trend in randomized placebo-controlled clinical
mine treatment were generally maintained within the mod- trials in AD that has not been addressed. Specifically, the
erate stage of AD severity, according to MMSE scoring, for change on the ADAS-cog is lessening in the placebo groups
up to five years after treatment initiation. Patients receiving with over time. Old studies report 7.0 ⫾ 7.8 points lost on the
rivastigmine for five years declined on average by 1.7 ADAS-Cog in placebo groups [82,83]. Newer studies have
MMSE points per year, which is smaller than the cognitive report less decline in the placebo groups (approximately five
decline predicted using both the basic model of “untreated” points per year) [84 – 87]. The rate of decline has clearly
decline and the more conservative baseline-dependent diminished. Some of this trend might be accounted for by the
model at three, four, and five years. Decline in the treated fact that randomized clinical trials a decade ago did not include
patients was also smaller than the decline reported for un- subjects taking ChEIs, but current studies mostly include those
treated patients in the literature (two to four points) [75]. subjects.
Grossberg et al [76] reports that AD patients taking
rivastigmine for two years had less decline on cognitive
6. An Opposing View: There is no Disease-Modifying
measures compared with expected decline in a projected
Effect
placebo group. A meta-analysis was performed recently
using data available for patients remaining on rivastigmine Recent data have asserted that the ChEIs may have only
treatment for up to five years [77]. The mean baseline symptomatic effects. Recently completed MCI studies for
122 M.N. Sabbagh et al. / Alzheimer’s & Dementia 2 (2006) 118 –125

donepezil of three years’ duration, rivastigmine for two 8. Conclusions

years’ duration, and galantamine for two years’ duration,
failed to show any effect for these drugs versus placebo on There is considerable preclinical, radiologic, and clin-
conversion to AD by the end of the trials. However, some ical evidence that cholinergic drugs such as ChEIs may
benefits were seen with donepezil at six, 12, and 18 months have disease-modifying effects. The linkage of musca-
[88]. These data suggest a pattern consistent with symptom- rinic receptor activation to ␤APP processing suggests
atic effects that may fade during long-term therapy. Simi- that cholinergic pharmacotherapy may influence amyloid
larly, a multiyear donepezil study in Britain, after three deposition, one of the fundamental processes thought to
years, failed to show significant benefits in slowing func- underlie disease progression. By reducing brain, CSF, or
tional decline or delaying nursing home placement [89]. plasma A␤ concentrations, cholinergic agents may be
This study suffered from poor design because of several able to slow 〈␤ deposition and hence retard disease
washouts, the lack of statistical power, and the very small progression. Because A␤ deposition begins as much as
sample size at the end of the study. This study did not 20 years before the clinical expression of AD and may
contribute much in understanding the long-term effects of already be quite prominent at the time of clinical diag-
ChEIs. nosis, the greatest possible effects of such therapy might
Despite the robust evidence for anti-amyloid effects of be achieved if it were instituted at very early presymp-
the muscarinic agonists, these drugs have not fared will in tomatic stages of disease.
randomized clinical trials. Trials of Lu 25-109 and xanome- These findings must be interpreted with caution for a
line have shown modest nonsignificant clinical benefit and few reasons. First, much of the preclinical data that
have been hampered by significant gastrointestinal side ef- observe an effect on APP are from muscarinic agonists
fects. These drugs have not been able to achieve approval and not ChEIs. Some of the preclinical data with ChEIs
despite several years of investigation. that do show significant A␤-lowering effects across dif-
ferent models have tested tacrine and phenserine, drugs
that may work by noncholinergic mechanisms such as
inhibition of the promoter for the APP gene. Muscarinic
7. Radiologic Evidence of ChEI’s Demonstrating agonists such as xanomeline and Lu25-109 have not
Disease-Modifying Effects succeeded in clinical development because of marginal
In a randomized clinical trial of 67 subjects with proba- cognitive effects and significant adverse cholinergic side
ble AD investigating hippocampal volume on volumetric effects. However, all of the clinical trials completed with
magnetic resonance imaging (vMRI), subjects were ex- muscarinic agonists were powered for symptomatic ef-
posed for 24 weeks to donepezil or placebo. After six fect and were not sufficiently long to ascertain whether
months, the placebo group had 8% loss in hippocampus, but there was a disease-modifying effect. Second, the ChEI
donepezil had only 0.5% loss [90]. A recent study by Hashi- long-term clinical studies that are very provocative for
moto et al [91] measured hippocampal vMRI for 12 months the possibility of a disease-modifying effect are con-
in a side-by-side comparison (donepezil vs no treatment). founded by significant and potentially selective attrition
They showed that treatment with donepezil reduced volume from the start to the end of the study. Long-term studies
loss by 24%. This suggests cholinomimetic therapy may that posit a disease-modifying effect support the obser-
delay neuronal or synaptic loss, the presumed microscopic vation of delay to long-term care placement seen with
substrates for macroscopic volume loss. tacrine and donepezil [96,97]. Another study found that
Long-term tacrine treatment has been shown to restore AD subjects that are long-term users of ChEIs were 2.5
nACh receptors as noted by positron-emission tomography times more likely to be classified with a slowly progress-
(PET). In a double-blind, parallel treatment fluoro-deoxy- ing disease course and a have a lower risk of being
glucose (FDG)-PET study, Tune et al [92] noted that done- admitted to a skilled nursing facility compared with AD
pezil-treated patients had no decrement in glucose metabo- subjects not taking ChEIs [98]. In that cohort, 50% of
lism, but the placebo group had decreased glucose subjects not taking ChEIs were in an skilled nursing
metabolism. Studies of rivastigmine showed similar results facility (SNF) compared with 11% [95]. In practice, the
suggesting little or no decline in glucose metabolism at six case for long-term disease-modifying effect of ChEIs
months [93] and at one year [94]. Although no long-term would be significantly bolstered by long-term population-
PET data have been published on AD subjects treated with based studies that track care of AD subjects (like Medi-
galantamine, there is recent evidence that behavioral and care or HMO databases). Ultimately, given the simulta-
cognitive improvements in response to galantamine in AD neous symptomatic effects of these medications, only a
subjects results in increased FDG-PET activity in cingulate long-term double-blind, placebo-controlled trial of at
and putamen [95]. These data suggest activation of brain least two years’ duration in patients with mild to moder-
regions beyond memory circuitry, and these changes are ate stage disease could definitely establish an effect for
sustained over time. ChEIs on delaying disease progression. The development
 M.N. Sabbagh et al. / Alzheimer’s & Dementia 2 (2006) 118 –125 123

of extensive A␤ deposition at early stages of AD sug- effects on cholinergic neurotransmission in rat cerebral cortex and
gests, however, that the most marked disease-altering hippocampus. Brain Res Bull 1995;39:371– 81.
[14] Lin L, LeBlanc CJ, Deacon TW, Isacson O. Chronic cognitive deficits
effects would be observed in a primary prevention trial. and amyloid precursor protein elevation after selective immunotoxin
lesions of the basal forebrain cholinergic system. Neuroreport 1998;
Acknowledgements 9:547–52.
[15] Beach TG, Walker D, Sue L, et al. Immunotoxin lesion of the
The authors thank Bonnie Tigner for the preparation of cholinergic nucleus basalis causes Ab deposition: towards a physio-
this manuscript. Supported by: NIA P30s AG 019610, AG logic animal model of Alzheimer’s disease. Curr Med Chem 2003;
3:57–75.
10133 and the Sun Health Research Institute. Dr Sabbagh [16] Roher AE, Kuo YM, Potter PE, et al. Cortical cholinergic denervation
receives grant support and honoraria from Pfizer and Eisai. elicits vascular A beta deposition. Ann NY Acad Sci. 2000;903:366 –
He is on the speaker’s bureau for Ortho-McNeil Neurolog- 73.
ical. Dr Farlow consults for and receives grants from No- [17] Beach TG, Walker DG, Roher AE, Potter PE. Anti-Amyloidogenic
vartis. He is on the speaker’s bureau for Pfizer. Dr Relkin is activity of cholinergic agents. Drug Dev Res 2002;56:242–7.
[18] Beach TG. Muscarinic agonists as preventative therapy for Alzhei-
on the speaker’s bureau for Eisai. Dr Beach has no conflict
mer’s disease, Curr Opin Investig Drugs 2002;3:1633–36.
of interest to report. [19] Farias GG, Godoy JA, Hernandez F, Avila J, Fisher A, Inestrosa NC.
M1 muscarinic receptor activation protects neurons from beta-amy-
loid toxicity: a role for Wnt signaling pathway. Neurobiol Dis 2004;
References
17:337– 48.
[20] Holschneider DP, Waite JJ, Li MG, Scremin OU. Decreased cortical
[1] Perry EK, Johnson M, Kerwin JM, et al. Convergent cholinergic blood flow after intraventricular administration of the 192 IgG-sa-
activities in aging and Alzheimer’s disease. Neurobiol Aging 1992; porin immunotoxin. Soc Neurosci Abs 1997;23:1575.
13:393– 400.
[21] Buxbaum JD, Oishi M, Chen HI, et al. Cholinergic agonists and
[2] Katzman R, Terry R, DeTeresa R, et al. Clinical, pathological, and
interleukin 1 regulate processing and secretion of the Alzheimer
neurochemical changes in dementia: a subgroup with preserved men-
beta/A4 amyloid protein precursor. Proc Natl Acad Sci U S A 1992;
tal status and numerous neocortical plaques Ann Neurol 1988;23:
89:10075– 8.
138 – 44.
[22] Nitsch RM, Slack BE, Wurtman RJ, Growdon JH. Release of Alz-
[3] Beach TG, Honer WG, Hughes LH. Cholinergic fibre loss associated
heimer amyloid precursor derivatives stimulated by activation of
with diffuse plaques in the non- demented elderly: the preclinical
muscarinic acetylcholine receptors, Science 1992;28:304 –7.
stage of Alzheimer’s disease? Acta Neuropathol (Berl) 1997;93:146 –
[23] Hung AY, Haass C. Nitsch RM, et al. Activation of protein kinase C
53.
inhibits cellular production of the amyloid beta-protein. J Biol Chem
[4] Beach TG, Potter PE, Kuo YM, et al. Cholinergic deafferentation of
1993;268:22959 – 62.
the rabbit cortex: a new animal model of Abeta deposition. Neurosci
[24] Bymaster FP, Wong DT, Mitch CH, et al. Neurochemical effects of
Lett 2000;283:9 –12.
the M1 muscarinic agonist xanomeline (LY246708/NNC11-0232).
[5] Potter P, Pandya Y, et al. Cortical cholinergic deficit is associated
J Pharmacol Exp Ther 1994;269:282–9.
with plaque development at preclinical stages of Alzheimer’s disease.
Neurobiol Aging 2004;25(S2):79. [25] Nitsch RM, Growdon JH. Role of neurotransmission in the regulation
[6] Funato H, Yoshimura M, Kusui K, et al. Quantitation of amyloid of amyloid beta-protein precursor processing. Biochem Pharmacol
beta-protein (A beta) in the cortex during aging and in Alzheimer’s 1994;47:1275– 84.
disease, Am. J. Pathol 1998;152:1633– 40. [26] Haring R, Gurwitz D, Barg J, et al. Amyloid precursor protein
[7] Davies L, Wolska B, Hilbich C, et al. A4 amyloid protein deposition secretion via muscarinic receptors: reduced desensitization using the
and the diagnosis of Alzheimer’s disease: prevalence in aged brains M1-selective agonist AF102B. Biochem Biophys Res Commun 1994;
determined by immunocytochemistry compared with conventional 203:652– 8.
neuropathologic techniques. Neurology 1988;38:1688 –1693. [27] Haring R, Gurwitz D, Barg J, et al. NGF promotes amyloid precursor
[8] Inestrosa NC, Alvarez A, Perez CA, et al. Acetycholinesterase accel- protein secretion via muscarinic receptor activation. Biochem Bio-
erates assembly of amyloid-beta-peptides into Alzheimer’s fibrils: phys Res Commun 1995;213:15–23.
possible role of the peripheral site of the enzyme. Neuron 1996;16(4): [28] Eckols K, Bymaster FP, Mitch CH, Shannon HE, Ward JS, DeLapp
881–91. NW. The muscarinic M1 agonist xanomeline increases soluble amy-
[9] Davis KL, Mohs RC, Marin D, et al. Cholinergic markers in elderly loid precursor protein release from Chinese hamster ovary-m1 cells.
patients with early signs of Alzheimer disease. JAMA 1999;281: Life Sci 1995;57:1183–90.
1401– 6. [29] Wolf BA, Wertkin AM, Jolly YC, et al. Muscarinic regulation of
[10] DeKosky ST, Ikonomovic MD, Styren SD, et al. Upregulation of Alzheimer’s disease amyloid precursor protein secretion and amyloid
choline acetyltransferase activity in hippocampus and frontal cortex beta-protein production in human neuronal NT2N cells. J Biol Chem
of elderly subjects with mild cognitive impairment, Ann Neurol 1995;270:4916 –22.
2002;51:145–55. [30] Muller DM, Mendla K, Farber SA, Nitsch RM. Muscarinic M1
[11] Wallace WC, Bragin V, Robakis NK, et al. Increased biosynthesis of receptor agonists increase the secretion of the amyloid precursor
Alzheimer amyloid precursor protein in the cerebral cortex of rats protein ectodomain. Life Sci 1997;60:985–91.
with lesions of the nucleus basalis of Meynert. Brain Res Mol Brain [31] Bymaster FP, Carter PA, Peters SC, et al. Compared to other mus-
Res 1991;10:173– 8. carinic agents on stimulation of phosphoinositide hydrolysis in vivo
[12] Wallace W, Ahlers ST, Gotlib J, et al. Amyloid precursor protein in and other cholinomimetic effects. Brain Res 1998;795:179 –90.
the cerebral cortex is rapidly and persistently induced by loss of [32] DeLapp N, Wu S, Belagaje R, et al. Effects of the M1 agonist
subcortical innervation. Proc Natl Acad Sci U S A 1993;90:8712– 6. xanomeline on processing of human beta-amyloid precursor protein
[13] Rossner S, Schliebs R, Hartig W, Bigl V. 192IGG-saporin-induced (FAD, Swedish mutant) transfected into Chinese hamster ovary-m1
selective lesion of cholinergic basal forebrain system: neurochemical cells. Biochem Biophys Res Commun 1998;244:156 – 60.
124 M.N. Sabbagh et al. / Alzheimer’s & Dementia 2 (2006) 118 –125

[33] Fisher A. Therapeutic strategies in Alzheimer’s disease: M1 musca- [52] Basun H, Nilsberth C, Eckman C, Lannfelt L, Younkin S. Plasma
rinic agonists. Jpn J Pharmacol 2000;84:101–12. levels of Abeta42 and Abeta40 in Alzheimer patients during treat-
[34] Hock C, Golombowski S, Muller-Spahn F, et al. Cerebrospinal fluid ment with the acetylcholinesterase inhibitor tacrine. Dement Geriatr
levels of amyloid precursor protein and amyloid beta-peptide in Cogn Disord 2002;14:156 – 60.
Alzheimer’s disease and major depression - inverse correlation with [53] Lahiri DK, Farlow MR, Sambamurti K. The secretion of amyloid
dementia severity. Eur Neurol 1998;39:111– 8. beta-peptide is inhibited the tacrine-treated human neuroblastoma
[35] Muller D, Wiegmann H, Langer U, Moltzen-Lenz S, Nitsch RM. Lu cells. Brain Res Mol Brain Res 1998;2(2):131– 40.
25-109, a combined m1 agonist and m2 antagonist, modulates regu- [54] Zimmerman M, Gardoni G, Marcello E, Colciaghi E, et al. Acetyl-
lated processing of the amyloid precursor protein of Alzheimer’s cholinesterase inhibitors increase ADAM10 activity by promoting its
disease. J Neural Transm 1998;105:1029 – 43. trafficking in neuroblastoma cell lines. J Neurochem 2004;90(6):
[36] Savage MJ, Trusko SP, Howland DS, et al. Turnover of amyloid 1489 –99.
beta-protein in mouse brain and acute reduction of its level by [55] Boronni B, Colchiaghi F, Pastorino L, Pettenati C, et al. Amyloid
phorbol ester. J Neurosci 1998;18:1743–52. precursor protein in platelets of patients with Alzheimer’s disease:
[37] Buxbaum JD, Koo EH, Greengard P. Protein phosphorylation inhibits effect of acetycholinesterase inhibitor treatment. Arch Neurol 2001;
production of Alzheimer amyloid beta/A4 peptide. Proc Natl Acad 58(3):442– 6.
Sci U S A 1993;90:9195– 8. [56] Borroni B, Colchiaghi F, Pastorino L, Archetti S, et al. ApoE geno-
[38] Slack B, Nitsch R, Livneh E, et al. Regulation by phorbol esters of type influences the biological effect at donepezil on APP metabolism
amyloid precursor protein release from Swiss 3T3 fibroblasts over- in Alzheimer disease: evidence from a peripheral model. Eur Neuro-
expressing protein kinase C alpha. J Bio Chem 1993;268:21097–101. psychopharmacol 2002;12(3):195–200.
[39] Greig NH, Sambamurti K, Yu QS, Brossi A, Bruinsma GB, Lahiri [57] Liu HC, Chi CW, Ko SY, Want HC, et al. Cholinesterase inhibitor
DK. An overview of phenserine tartrate, a novel acetylcholinesterase affects the amyloid precursor protein isoforms in patients with Alz-
inhibitor for the treatment of Alzheimer’s disease. Curr Alzheimer heimer’s disease. Dementia Geriatr Cogn Disord 2005;19(5-6):
Res 2005;2:281–90. 345– 8.
[40] Shaw KT, Utsuki T, Rogers J, Yu QS, Sambamurti K, Brossi A, et al. [58] Hellstrom-Lindahl E, Court J, Keverne J, et al. Nicotine reduces A
Phenserine regulates translation of beta -amyloid precursor protein beta in the brain and cerebral vessels of APPsw mice. Eur J Neurosci
mRNA by a putative interleukin-1 responsive element, a target for 2004;19:2703–10.
drug development. Proc Natl Acad Sci U S A 2001;98(13):7605–10. [59] Nordberg A, Hellstrom-Lindahl E, Lee M, et al. Chronic nicotine
[41] Lahiri DK, Farlow MR. Differential effect of tacrine and physostig- treatment reduces beta-amyloidosis in the brain of a mouse model of
mine on the secretion of the beta-amyloid precursor protein in cell Alzheimer’s disease (APPsw) J Neurochem 2002;81:655– 8.
lines. Mol Neurosci 1996;7(1):41–9. [60] Arias E, Gallego-Sandin S, Villaroya M, Garcia AG, Lopez MG.
[42] Farber SA, Nitsch RM, Schulz JG, Wurtman RJ. Regulated secretion Unequal neuroprotection afforded by the acetylcholinesterase inhib-
of beta-amyloid precursor protein in rat brain, J Neurosci 1995;15: itors galantamine, donepezil, and rivastigmine in SH-SY5Y neuro-
7442–51. blastoma cells: roll of nicotinic receptors. J Pharmacol Exp Ther
[43] Mori F, Lai CC, Fusi F, Giacobini E. Cholinesterase inhibitors in- 2005;315 (3):1346 –53.
crease secretion of APPs in rat brain cortex, Neuroreport 1995;6: [61] Arias E, Ales E, Gabilan NH, Cano-Abad MD, Villaroya M, Garcia
633– 6. AG, et al. Galantamine prevents apoptosis induced by beta-amyloid
[44] Chong YH, Suh YH. Amyloidogenic processing of Alzheimer’s amy- and thapsigargin: involvement of nicotinic acetylcholine receptors.
loid precursor protein in vitro and its modulation by metal ions and Neuropharmacology 2004;46(1):103–14.
tacrine. Life Sci 1996;59(7):545–7. [62] Kihara T, Sawada H, Nakmizo T, Kanki R, Yamashita H, Maelicke
[45] Racchi M, Sironi M, Caprera A, Konig G, Govoni S. Short- and A, et al. Galantamine modulate nicotinic receptor and blocks Abeta-
long-term effect of acetylcholinesterase inhibition on the expression enhanced glutamate toxicity. Biochem Biophys Res Commun 2004;
and metabolism of the amyloid precursor protein. Mol Psychiatry 325(3):976 – 82.
2001;6(5):520 – 8. [63] Bullock R, Dengiz A. Cognitive performance in patients with Alz-
[46] Pittel Z, Heldman E, Barg J, Haring R, Fisher A. Muscarinic control heimer’s disease receiving cholinesterase inhibitors for up to 5 years.
of amyloid precursor protein secretion in rat cerebral cortex and Int J Clin Pract 2005;Jul; 59(7):817–22.
cerebellum. Brain Res 1996;742:299 –304. [64] Stern RG, Mohs RC, Davidson M, et al. A longitudinal study of
[47] Haroutunian V, Greig N, Pei XF, et al. Pharmacological modulation Alzheimer’s disease: measurement, rate, and predictors of cognitive
of Alzheimer’s beta-amyloid precursor protein levels in the CSF of deterioration. Am J Psychiatry 1994;151(3):390 – 6.
rats with forebrain cholinergic system lesions. Brain Res Mol Brain [65] Doraiswamy PM, Kaiser L, Bieber F, Garman RL. The Alzheimer’s
Res 1997;46:161– 8. Disease Assessment Scale: evaluation of psychometric properties and
[48] Beach TG, KuoY ??, Schwab C, Walker DG, Roher AE. Reduction patterns of cognitive decline in multicenter clinical trials of mild to
of cortical amyloid beta levels in guinea pig brain after systemic moderate Alzheimer’s disease. Alzheimer Dis Assoc Disord 2001;
administration of physostigmine. Neurosci Lett 2001;310:21– 4. 15(4):174 – 83.
[49] Beach TG, Walker DG, Potter PE, Sue LI, Fisher A. Reduction of [66] Doody RS, Dunn JK, Clark CM, et al. Chronic donepezil treatment is
cerebrospinal fluid amyloid beta after systemic administration of M1 associated with slowed cognitive decline in Alzheimer’s disease.
muscarinic agonists. Brain Res 2001;905:220 –3. Dement Geriatr Cogn Disord 2001;12(4):295–300.
[50] Nitsch RM, Deng M, Tennis M, Schoenfeld D, Growdon JH. The [67] Winblad B, Engedal K, Soininen H, et al. A 1-year, randomized,
selective muscarinic M1 agonist AF102B decreases levels of total placebo-controlled study of donepezil in patients with mild to mod-
Abeta in cerebrospinal fluid of patients with Alzheimer’s disease, erate AD. Neurology. 2001;Aug 14;57(3):489 –95.
Ann Neurol 2000;48:913– 8. [68] Matthews HP, Korbey J, Wilkinson DG, Rowden J. Donepezil in
[51] Hock C, Maddalena A, Raschig A, et al. Treatment with the selective Alzheimer’s disease: eighteen months from Southampton Memory
muscarinic m1 agonist talsaclidine decreases cerebrospinal fluid lev- Clinic. Int J Geriatric Psychiatry 2000;15:713–20.
els of A beta 42 in patients with Alzheimer’s disease. Amyloid [69] Rogers SL, Friedhoff LT. Long-term efficacy and safety of donepezil
2003;10:1– 6. in the treatment of Alzheimer’s disease: interim analysis of the results
 M.N. Sabbagh et al. / Alzheimer’s & Dementia 2 (2006) 118 –125 125

of a US multicentre open label extension study. Eur Neuro- [85] Aisen PS, Schafer KA, Grundman M, Pfeiffer E, et al. Effects of
psychopharmacol 1998;8:67–75. rofecoxib or naproxen vs placebo on Alzheimer disease progression:
[70] Doody R, Geldmacher D, Gordon B, et al. Open-label, multicenter, a randomized controlled trial. JAMA 2003;289(21);2819 –26.
phase 3 extension study of the safety and efficacy of donepezil in [86] Reines SA, Block GA, Morris JC, et al. Rofecoxib: no effect on
patients with Alzheimer disease. Arch Neurol 2001;58:427–33. Alzheimer’s disease in a 1-year, randomized, blinded, controlled
[71] Rogers SL, Doody RS, Pratt RD, Ieni JR. Long-term efficacy and study. Neurology 2004;62(1):66 –71.
safety of donepezil in the treatment of Alzheimer’s disease: final [87] Gilman S, Koller M, Black RS, Jenkins L, et al Clinical effects of
analysis of a US multicentre open label extension study. Eur Neuro- Abeta immunization (AN1792) 2005;64(9):1553– 62.
psychopharmacol 2000;10:195–203. [88] Petersen RC, Thomas RG, Grundman M, et al. Alzheimer’s Disease
[72] Farlow M, Potkin S, Koumaras B, Veach J, Mirski D. Analysis of Cooperative Study Group. Vitamin E and donepezil for the treatment
outcome in retrieved dropout patients in a rivastigmine vs placebo, of mild cognitive impairment. N Engl J Med 2005;352(23):2379 – 88.
26-week, Alzheimer disease trial. Arch Neurol 2003;60(6):843– 8. Epub 2005 Apr 13.
[73] Leber P. Observations and suggestions on antidementia drug devel- [89] Courtney C, Farrell D, Gray R, Hills R, et al. Long-term donepezil
opment. Alzheimer Dis Assoc Disord 1996;10(suppl1):31–5. treatment in 565 patients with Alzheimer’s disease AD 2000: ran-
[74] Small GW, Kaufer D, Mendiondo MS, Quarg P, Spiegel R. Cognitive domised double-blind trial. Lancet 2004;363(9427):2105–15.
performance in Alzheimer’s disease patients receiving rivastigmine [90] Krishnan KRR, Charles HC, Doraiswamy PM. et al. Randomized,
for up to 5 years. Int J Clin Pract 2005;59(4):473–7. placebo controlled trial of the effects of donepezil on neuronal mark-
[75] Rainer M, Mucke HAM. Int J Geriatr Psychopharmacol 1998;1:197– ers and hippocampal volumes in Alzheimer’s disease. Am J Psychi-
201. atry 2003;160:2003–11.
[76] Grossberg C, Irwin P, Satlin A, Mesenbrink P, Spiegel R. Rivastig- [91] Hashimoto M, Kazui H, Matsumoto K, Nakano Y, Yasuda M, Mori
mine in Alzheimer disease: efficacy over two years. Am J Geriatr E. Does donepezil treatment slow the progression of hippocampal
Psychiatry 2004;12(4):420 –31. atrophy in patients with Alzheimer’s disease? Am J Psychiatry 2005;
[77] Small GW, Kaufer D, Quarg P, Spiegel R. Poster presented at the 9th 162(4):676 – 82.
ICADRD. Neurobiol Aging 2004;25(S2):79. [92] Tune LE, Hoffman, JM, Tisco PJ, et al. Donepezil HCL maintains
[78] Tariot PN, Kershaw P. Galantamine demonstrates cognitive benefits global glucose activity in patients with Alzheimer’s disease; results of
in patients with AD for at least 18 months. Neurology 2002;58: a 24 weeks study. J Cereb Blood Flow Metab 1999;19(suppl 1):5838.
A277– 8. [93] Potkin SG, Anand R, Fleming K, Alva G, et al Brain metabolic and
[79] Pirttila T, Wilcock G, Truyen L, Damaraju CV. Long-term efficacy clinical effects of rivastigmine in Alzheimer’s disease. Int J Neuro-
and safety of galantamine in patients with mild-to-moderate Alzhei- psychopharmacol 2001;4(3):223–30.
mer’s disease: multicenter trial. Eur J Neurol 2004;Nov;11(11):734 – [94] Stefanova E, Wall A, Almkvist O, et al. Longitudinal PET evaluation
41. of cerebral glucose metabolism in rivastigmine treated patients with
[80] Raskind MA, Peskind ER, Truyen L, Kershaw P, Damaraju CV. The mild Alzheimer’s disease. J Neural Transmission 2005;July 29[epub
cognitive benefits of galantamine are sustained for at least 36 months. ahead of print].
Arch Neurol 2004;61:252– 6. [95] Mega MS, Dinov ID, Porter V, Chow G, et al. Metabolic patterns
[81] Farlow MR, Lilly ML, for the ENA 713 B352 Study Group. Riv- associated with the clinical response to galantamine therapy: a flu-
astigmine: an open-label, observational study of safety and effective- orodeoxyglucose f18 positron emission tomographic study. Arch
ness in treating patients with Alzheimer’s disease for up to 5 years. Neurol 2005;62(5);721– 8.
BMC Geriatrics 2005;5(1):3. [96] Knopman D, Schneider L, Davis K, et al. Long-term tacrine (Cognex)
[82] Thal LJ, Carta A, Clarke WR, Ferris SH, et al. A 1-year multicenter, treatment: effects on nursing home placement and mortality, Tactrine
placebo-controlled study of acetyl-l-carnitine in patients with Alzhei- Study Group. Neurology 1996;July:47(1):166 –177.
mer’s disease. Neurology 1996;47(3):705–11. [97] Geldmacher D, Provenzanok G, McRae T, et al. Donepezil is asso-
[83] Thal LJ, Calvani M, Amato A, Carta A. A 1-year controlled trial of ciated with delayed nursing home placement in patients with Alzhei-
acetyl-l-carnitine in early-onset AD. Neurology 2000;55(6):805–10. mer’s disease. J Am Ger Soc 2003;51:937– 44.
[84] Aisen PS, Davis KL, Berg JD, Schafer K, et al. A randomized [98] Lopez OL, Becker JT, Saxton J, Sweet RA, Klunk W, DeKosky ST.
controlled trial of prednisone in Alzheimer’s disease. Alzheimer’s Alteration of a clinically meaningful outcome in the natural history of
Disease Cooperative Study. Neurology 2000;54(3):588 –93. Alzheimer’s disease by cholinesterase inhibition. 2005;53(1):83–7.