REVIEW

The cholinergic hypothesis of Alzheimers disease:a review of progress

Paul T Francis, Alan M Palmer, Michael Snape, Gordon K Wilcock

AbstractAlzheimers disease is one of the mostcommon causes of mental deterioration inelderly people, accounting for around 50%-60% of the overall cases of dementia amongpersons over 65 years of age. The past twodecades have witnessed a considerableresearch eVort directed towards discover-ing the cause of Alzheimers disease withthe ultimate hope of developing safe andeVective pharmacological treatments. Thisarticle examines the existing scientificapplicability of the original cholinergichypothesis of Alzheimers disease by de-scribing the biochemical and histopatho-logical changes of neurotransmittermarkers that occur in the brains of patientswith Alzheimers disease both at postmor-tem and neurosurgical cerebral biopsy andthe behavioural consequences of cholino-mimetic drugs and cholinergic lesions.Such studies have resulted in the discoveryof an association between a decline inlearning and memory, and a deficit in exci-tatory amino acid (EAA) neurotransmis-sion, together with important roles for thecholinergic system in attentional process-ing and as a modulator of EAA neurotrans-mission. Accordingly, although there ispresently no cure for Alzheimers dis-ease, a large number of potential therapeu-tic interventions have emerged that aredesigned to correct loss of presynapticcholinergic function. A few of these com-pounds have confirmed eYcacy in delayingthe deterioration of symptoms ofAlzheimers disease, a valuable treatmenttarget considering the progressive natureof the disease. Indeed, three compoundshave received European approval for thetreatment of the cognitive symptoms ofAlzheimers disease, first tacrine and morerecently, donepezil and rivastigmine, all ofwhich are cholinesterase inhibitors.(J Neurol Neurosurg Psychiatry 1999;66:137147)

Keywords: acetylcholine; Alzheimers disease; cholinest-erase inhibitors; treatment

Alzheimers disease aVects an estimated 15million people worldwide and is the leading

cause of dementia in elderly people. With theproportion of elderly people in the populationincreasing steadily, the burden of the disease,both to carers and national economies, isexpected to become substantially greater overthe next 2 to 3 decades.

Alzheimers disease is a progressive neurode-generative disorder with a mean duration ofaround 8.5 years between onset of clinicalsymptoms and death. Brain regions that areassociated with higher mental functions, par-ticularly the neocortex and hippocampus, arethose most aVected by the characteristicpathology of Alzheimers disease. This includesthe extracellular deposits of -amyloid (derivedfrom amyloid precursor protein; APP) in senileplaques, intracellular formation of neurofibril-lary tangles (containing an abnormally phos-phorylated form of a microtubule associatedprotein, tau), and the loss of neuronal synapsesand pyramidal neurons. These changes resultin the development of the typical symptomol-ogy of Alzheimers disease characterised bygross and progressive impairments of cognitivefunction and often accompanied by behav-ioural disturbances such as aggression, depres-sion, and wandering. Carers find these featuresthe most diYcult to cope with and they oftenlead to the need for institutionalisation of thepatient.1

The systematic biochemical investigation ofthe brains of patients with Alzheimers diseasebegan in the late 1960s and early 1970s. Thehope was that a clearly defined neurochemicalabnormality would be identified, providing thebasis for the development of rational therapeu-tic interventions analogous to levodopa treat-ment of Parkinsons disease. Support for thisperspective came in the mid-1970s withreports of substantial neocortical deficits in theenzyme responsible for the synthesis of acetyl-choline (ACh), choline acetyltransferase(ChAT).24 Subsequent discoveries of reducedcholine uptake,5 ACh release6 and loss ofcholinergic perikarya from the nucleus basalisof Meynert7 confirmed a substantial presynap-tic cholinergic deficit.

These studies, together with the emergingrole of ACh in learning and memory,8 led to thecholinergic hypothesis of Alzheimers disease(figure A). Thus it was proposed that degenera-tion of cholinergic neurons in the basal

J Neurol Neurosurg Psychiatry 1999;66:137147 137

Dementia ResearchLaboratory,Neuroscience ResearchCentre, Guys, Kingsand St ThomasSchools of BiomedicalSciences, KingsCollege, London, SE19RT, UKP Francis

Cerebrus, OakdeneCourt, 613 ReadingRoad, Winnersh,Wokingham, RG415UA, UKA PalmerM Snape

Department of Care ofthe Elderly, FrenchayHospital, Bristol, BS162EW, UKG Wilcock

Correspondence to:Dr Paul T Francis, DementiaResearch Laboratory,Division of BiomolecularSciences, Guys, Kings andSt Thomas Schools ofBiomedical Sciences, KingsCollege, St Thomas Street,London SE1 9RT, UK.Telephone 0044 171 9552611; fax and answer phone0044 171 955 2600; emailp.francis@umds.ac.uk

Received 11 June and inrevised form20 October 1998Accepted 30 October 1998

forebrain and the associated loss of cholinergicneurotransmission in the cerebral cortex andother areas contributed significantly to thedeterioration in cognitive function seen inpatients with Alzheimers disease.9

Over the 20 years since the origins of thecholinergic hypothesis, data from numerousstudies have challenged its veracity as an expla-nation for the syndrome of dementia inAlzheimers disease. Thus, this review attemptsto re-evaluate the cholinergic hypothesis in thefollowing ways:

(1) Setting the original findings of reducedcholinergic neurotransmission in the context ofchanges in other neurotransmitter systems, aclear understanding of the behavioural role of

the cholinergic system, and a more detailedunderstanding of the molecular pathology ofthe disease.

(2) Charting the preclinical and clinicaldevelopment of cholinomimetic drugs for thesymptomatic treatment of Alzheimers disease,focusing on the first generation and secondgeneration cholinesterase inhibitors currentlyavailable.

Neurochemical and histopathologicalchanges in cholinergic andnon-cholinergic neurons in AlzheimersdiseaseAt postmortem, Alzheimers disease is charac-terised by neuronal loss and neurofibrillary

Schematic diagram of a neuron representing (A) alterations in neurotransmission in Alzheimers disease and (B) thehypothetical mode of action of AChE inhibitors. Key to figure (A): (1) reduced cortical cholinergic innervation; (2) reducedcorticocortical glutamatergic neurotransmission due to neuron or synapse loss; (3) reduced coupling of muscarinic M1receptors to second messenger system?; (4) shift of tau to the hyperphosphoryalted stateprecursor of neurofibrillary tangles;(5) reduced secretion of soluble APP; (6) increased production of -amyloid protein; (7) decreased glutamate production.*It is hypothesised that these changes give rise to the clinical symptoms of Alzheimers disease and contribute to the spreadof pathology.12 49 54 Key to figure B: (1) AChE inhibitors reduce the breakdown of endogenously released ACh, resulting ingreater activation of postsynaptic ACh receptors; hypothesised consequences: (2) reduced phosphorylation of tau; (3)secretion of sAPP returned towards normal; (4) reduced -amyloid production; (5) glutamatergic neurotransmission returnstowards normal, possibly due to activation of muscarinic and nicotinic receptors. ACh=acetylcholine; mAChR=AChmuscarinic receptor; APP=a myloid precursor protein; AChE=acetylcholinesterase; nAChR=ACh nicotinic receptor;Glu=glutamate.

1

AChAChE

nAChR

mAChR

Glu

tau tau- P

3

4

2

A Proposed neurochemical changes in Alzheimer's disease

Cortical pyramidal neuron

APP

sAPP 5

A 6

Glu 5

AChAChE

AChE-Inhibitors

Glu

tau tau- P

1

2

B Rectification of neurotransmission with cholinesterase inhibitors

Cortical pyramidal neuron

APP

sAPP 3

A 4

Glu 5

138 Francis, Palmer, Snape, et al

tangle formation in circumscribed regions ofthe neocortex and hippocampus, primarilyaVecting pyramidal neurons and theirsynapses.10 11 Neurotransmitter specific subcor-tical nuclei that project to the cortex are alsoaVected by neurodegenerative processes, in-cluding the cholinergic nucleus basalis of Mey-nert and medial septum, the serotonergic raphenuclei, and the noradrenergic locus coeruleus.

Biochemical investigations of biopsy tissuetaken from patients with Alzheimers disease3.5 years (on average) after the onset of symp-toms indicate that a selective neurotransmitterpathology occurs early in the course of thedisease.12 Specifically, presynaptic markers ofthe cholinergic system appear uniformly re-duced. This is exemplified by reductions inChAT activity and ACh synthesis which arestrongly correlated with the degree of cognitiveimpairment in patients with Alzheimersdisease.1215 Whereas serotonergic and somenoradrenergic markers are aVected, markersfor dopamine, -aminobutyric acid (GABA), orsomatostatin are not altered.12 When postmor-tem studies of Alzheimers disease brain areconsidered (typically representing a later stageof the disease) many more neurotransmittersystems are involved or are aVected to a greaterextent. These include GABA16 17 andsomatostatin18 19 and may indicate that corticalinterneurons, for which these are neurochemi-cal markers, are aVected later in the diseaseprocess. Based on postmortem studies, how-ever, changes in serotonergic neurotransmis-sion may be linked to the behavioural distur-bances of Alzheimers disease such asdepression, rather than cognitivedysfunction.1 20 21

On the basis of the above evidence, neocorti-cal cholinergic innervation is probably lost atan early stage of the disease, a conclusion sub-stantiated by evidence for similar changes inpatients that have displayed clinical symptomsfor less than 1 year.22 However, although theloss of cholinergic function is correlated withthe cognitive impairment in Alzheimers dis-ease, an association between two such indicesdoes not necessarily indicate a causal relation.Other indices also correlate with measures ofcognitive decline in Alzheimers disease, suchas loss of synapses and pyramidal cellperikarya.23 Moreover, a few patients withAlzheimers disease do not show large de-creases in ChAT activity, albeit that a smallreduction is found in the amygdala.24 Inaddition, patients with inherited olivopon-tocerebellar atrophy have diminished ChATactivity of a magnitude similar to that seen inAlzheimers disease in the absence of cognitivedeficits.25 Thus, although diminished ChATactivity is a necessary correlate of Alzheimersdisease, additional factors other than impairedcholinergic function are likely to participate inthe decline in cognitive function. Other studieshave demonstrated a reduction in the numberof nicotinic26 and muscarinic (M2) ACh recep-tors in Alzheimers disease brains, most ofwhich are considered to be located on presyn-aptic cholinergic terminals, but a relative pres-ervation of postsynaptic muscarinic (M1, M3)

receptors.27 However, there is some evidencefor a disruption of the coupling between themuscarinic M1 receptors, their G-proteins,and second messenger systems.28

In addition to cholinergic dysfunction, otherstrong correlates of dementia are the chemicaland histopathological markers of excitatoryamino acid (EAA) releasing cortical pyramidalneurons. These neurons, considered to con-tribute to normal cognitive function in theirown right, also seem to have a pivotal role incholinergic function as they arecholinoceptive.2932 Although neurochemicalstudies of EAA neurotransmission have failedto show profound or extensive alterations inEAA neuronal indices,12 this may be related tothe diYculty in distinguishing the transmitterpool of aspartate and glutamate from themetabolic pool. Nevertheless, glutamate con-centration was reduced by 14% in temporallobe biopsy samples of patients withAlzheimers disease. Greater reductions wereevident at postmortem in regions enriched withEAA nerve terminals.33 Uptake of D-aspartate,a putative marker of EAA nerve endings, is alsoreduced in many cortical areas in Alzheimersdisease brains.3436

Arguably, in vivo imaging studies of patientswith Alzheimers disease also support theinvolvement of pyramidal neurons in thedisease as the pattern of regional hypometabo-lism parallels neuronal loss/atrophy, tangle for-mation, and synapse loss.10 3739 Loss of corticalpyramidal neurons,23 40 41 synapse loss,40 andreduced glutamate concentration,17 togetherwith the formation of neurofibrillary tangles,42

all correlate with the severity of dementia.These findings indicate that pyramidal neuronsand their transmitter glutamate (and/or aspar-tate) play a part in the cognitive symptoms ofAlzheimers disease and may therefore repre-sent an additional therapeutic target. However,these neurons are cholinoceptive and it isreasonable to propose that one of the actions ofcholinomimetic drugs for the treatment ofAlzheimers disease is to increase the activity ofEAA neurons through muscarinic and nico-tinic receptors that are present on such cells.29

This is supported by electrophysiological stud-ies showing the excitatory actions of cholino-mimetic drugs in cortical pyramidal neuronsfrom both rats and humans30 31 and microdialy-sis studies in rats.32 Clearly, as a result ofcholinergic and other pyramidal neuronal loss,the profound reduction in EAA neurotransmis-sion will lead to pyramidal hypoactivity com-pounded by maintained levels of inhibition byGABAergic neurons. Consequently, it may behypothesised that in addition to the deleteriouseVects of neuronal loss and tangle formation,there is a change in the balance of neurotrans-mission in the Alzheimers disease brainfavouring lower neuronal activity.12 This maybe reflected in the hypometabolism in patientswith Alzheimers disease seen with imagingtechniques, although a component of this isalso likely to be due to neuronal atrophy.43

Likewise, it is of interest that regional cerebralblood flow may be increased in patients with

The cholinergic hypothesis of Alzheimers disease: a review 139

Alzheimers disease by cholinesterase (ChE)inhibitors such as physostigmine.44 45

CHOLINERGIC AND NON-CHOLINERGIC NEURONSAND ALZHEIMERS DISEASE NEUROPATHOLOGYThe discovery that rare mutations in the geneencoding for APP always led to Alzheimersdisease in family members carrying the defectresulted in the proposal of the amyloidcascade hypothesis of Alzheimers disease.46

Thus, the mismetabolism of APP leading toincreased production of -amyloid was pro-posed as the critical event in both familial andsporadic Alzheimers disease with otherchanges, tangles, neuron loss, synapse loss, andneurotransmission dysfunction, following as aconsequence. Cholinergic neurotransmissionmay be a specific target for -amyloid, as it hasbeen shown to reduce both choline uptake andACh release in vitro.47 48 It is of interest herethat disease related changes in the Alzheimersdisease brain are focused on pyramidal neuronsin that these cells are lost in the disease, subjectto tangle formation, represent a major sourceof APP (and hence, a site for its mismetabolismleading to increased -amyloid production)and are regulated by a neurotransmitter (ACh),aVected early in the disease. These neuronstherefore seem to have a central role in theclinical symptoms as well as in the pathophysi-ology of the disease. Observations in cell linesand primary neuronal cultures that the activa-tion of muscarinic, metabotropic glutamate,and other phospholipase C-linked receptorsfavours the non-amyloidogenic processing ofAPP49 suggests that compounds being devel-oped for symptomatic treatment may have aserendipitous eVect on the continuing emer-gence of pathology by reducing the productionof -amyloid. Furthermore, -amyloid neuro-toxicity is attenuated by muscarinic agonists.50

No data have yet been reported regarding thepotential beneficial eVects of cholinomimeticdrugs on either increasing APP or reducing-amyloid production in patients withAlzheimers disease. There is, however, someevidence for reductions in CSF fluid APP indepressed patients receiving drugs with anti-cholinergic side eVects.51 Clearly, long termstudies are called for to test this hypothesis inthe patient population. However, this may raiseethical problemsfor example, the need forserial lumbar puncture and the justification forgroups of patients to act as placebo controls.

Other studies have shown that the phospho-rylation of tau, thought to be an important stepin the formation of tangles (which occurpredominately in EAA cortical pyramidal neu-rons), may also be influenced by the phospholi-pase C second messenger system.52 Thus, aftermuscarinic cholinergic receptor stimulation,activation of protein kinase C may lead to theinactivation of a protein kinase (GSK-3) whichphosphorylates tau, in vitro, in a similarmanner to that found in Alzheimers disease.52

In support of this tenet, neuronal cells inculture transfected with M1 muscarinic recep-tors show reduced phosphorylation of tau aftertreatment with cholinergic agonists.53 There-fore, as a consequence of reduced cholinergic

activity, reduced activation of protein kinase Cmay lead to a higher level of activity of GSK-3and hence hyperphosphorylation of tau. Thus,if these neurotransmitter-protein interactionsoccur in the Alzheimers disease brain, it is notinconceivable that the changes in the balance ofneurotransmission in the Alzheimers diseasebrain may contribute to increased tau phos-phorylation and -amyloid production andhence neurodegeneration in selectively vulner-able regions. Furthermore, it is possible thatChE inhibitors may reduce the histopathologi-cal features of disease progression.

On the basis of recent studies of Alzheimersdisease, a glutamatergic hypothesis ofAlzheimers disease has been proposed as anauxiliary to the cholinergic hypothesis.12 54

Thus, the cholinergic hypothesis may berefined to include the idea that a major target ofcholinomimetic action is EAA pyramidal neu-rons, and that cholinergic hypofunction com-pounds the loss of EAA function. Togetherthese systems may be largely responsible for theneuropsychological deficits and may contrib-ute to the continuing emergence of pathologyin patients with Alzheimers disease. Thisrevised cholinergic hypothesis provides astronger case for the continued development ofcholinomimetic drugs for the symptomatictreatment of Alzheimers disease.

Behavioural consequences ofcholinomimetic drugs and cholinergiclesionsMany pharmacological studies have examinedthe eVect of cholinomimetic drugs and cholin-ergic receptor antagonists on learning andmemory tasks. The most commonly usedmodel is based on the finding that scopo-lamine, a muscarinic receptor antagonist,induces amnesia in young healthy subjectscomparable with that in old, untreatedsubjects.8 These deficits may be reversed byChE inhibitors. Compounds that reverse thesescopolamine induced deficits in experimentalanimals may be considered as potential drugsto treat cognitive impairment.

It is, however, diYcult to separate reliablythe eVects on learning and memory processesfrom eVects on other behavioural domains. Forexample, methylscopolamine (which does notcross the blood-brain barrier) is as active asscopolamine in several models of cognitivefunction,55 56 indicating that peripheral changesinduced by these compounds indirectly influ-ence performance in cognitive tasks. It is,therefore, very important to distinguish centralversus peripheral eVects of cholinmineticagents. Scopolamine induced impairment ofperformance may also be mediated by directeVects on sensorimotor function or motivationdeficits.56 57 Further, it is likely that thescopolamine induced impairment in the per-formance of both experimental animals andhumans in the delayed matching to positiontask (a commonly used test of cognitivefunction) is secondary to attentional deficitsthat are induced by the drug.58 59

Both hippocampal and cortical areas of thebrain receive major cholinergic input from

140 Francis, Palmer, Snape, et al

basal forebrain nuclei. Thus, the lesioning ofthese nuclei has been used to model choliner-gic denervation in Alzheimers disease and toestablish the behavioural consequences ofcholinergic deaVerentation. The most signifi-cant and consistent eVects of such cholinergiclesioning on learning and memory followlesioning of cholinergic pathways that lead tothe hippocampus.60 61 Initial studies used stere-otaxic injection of ibotenic acid to lesioncholinergic nuclei, and caused profound defi-cits in discrimination learning and memory.However, injection of the toxins quisqualic acidand -amino-3-hydroxy-5-methyl-4-isoxazole(AMPA) into the same site causes a greater lossof ChAT activity than ibotenate but only mar-ginal impairments in the same range of cogni-tive tasks.62 Thus, in addition to the establishedrole for ACh in learning and memory, there aredata to suggest that ACh also plays a criticalpart in attentional processing.6365 This issupported by a study showing that both tacrineand nicotine improve attentional functions inpatients with Alzheimers disease.66

Cholinomimetic therapy in AlzheimersdiseaseA prediction of the cholinergic hypothesis isthat drugs that potentiate central cholinergicfunction should improve cognition and per-haps even some of the behavioural problemsexperienced with Alzheimers disease. Thereare a number of approaches to the treatment ofthe cholinergic deficit in Alzheimers disease,most of which have initially focused on thereplacement of ACh precursors (choline orlecithin) but these agents failed to increasecentral cholinergic activity. Other studies haveinvestigated the use of ChE inhibitors thatreduce the hydrolysis of ACh (figure, B)forexample, physostigmine. More recent investi-gational compounds include specific M1 mus-

carinic or nicotinic agonists, M2 muscarinicantagonists, or improved second generationChE inhibitors (table).

Additional potential symptomatic therapeu-tic avenues relevant to the cholinergic hypoth-esis of Alzheimers disease have resulted fromthe rapid development in the understanding ofthe molecular pathology of the disease. Forexample, during the development of choliner-gic neurons in the basal forebrain, they expressfunctional nerve growth factor (NGF) recep-tors. In adult life, these neurons seem to remainresponsive to NGF. Consequently, intraven-tricular administration of NGF has beenshown to prevent the lesion induced loss ofcholinergic neuronal cell bodies and to acceler-ate the recovery of behavioural deficits inlearning.67 Another approach is the transplan-tation of ACh rich foetal tissue grafts, whichhas been shown to improve the cognitiveperformance of primates after excitotoxiclesions of cholinergic nuclei.68 Thus, althoughsuch approaches may provide additional futurepossibilities for the palliative treatment forAlzheimers disease, the use of ChE inhibitorsis the most well developed approach totreatment to date.

PRECLINICAL STUDIES OF CHOLINESTERASE

INHIBITORS

Although a variety of ChE inhibitors have beendeveloped as potential treatments forAlzheimers disease, their pharmacologicalactivities diVer. One of the most fundamentaldiVerences between them is in the mechanismof ChE inhibition. For example, enzyme kineticstudies have shown that tacrine, an acridinecompound, and donepezil, a novel piperidineclass agent, are mixed type reversible inhibi-tors of ChE. That is, these compounds inhibitChE via both non-competitive (by blockade ofthe deacetylation process) and ACh competi-

Cholinomimetic drugs in clinical development for Alzheimers disease including European registrations

Phase Drug name Company Mechanism of action

Registered Cognex (tacrine) Parke-Davis AChE inhibitorRegistered Aricept (donepezil) Eisai Inc/Pfizer Inc AChE inhibitorRegistered Exelon (rivastigmine) Novartis Inc AChE inhibitorAwaiting approval Metrifonate Bayer AG AChE inhibitorAwaiting approval Synapton (physostigmine) Forest Laboratories Inc AChE inhibitor3 SB-202026 SmithKline Beecham plc M1 agonist3 Amridin Nikken Chemicals Co Ltd AChE inhibitor3 Talsaclidine Boehringer Ingelheim Corp M1 agonist, nootropic agent3 Galantamine Janssen Ltd and Shire Ltd AChE inhibitor3 Montirelin Gruchenthal GmbH ACh release stimulator3 FKS-508 Nippon Shinyaku Co Ltd ACh modulator3 Nefiracetam Daijchi Sciyaku Co Ltd ACh receptor agonist3 AF-1025B (cevimeline) Snow Brand Milk Products Co Ltd M1agonist3 Eptastigmine Mediolanum Farmaceutical SpA AChE inhibitor3 Milameline Parke Davis and Co M1 agonist2 ABT-418 Abbot Laboratories nAChR agonist2 KA-672 Dr Wilmar Schwabe GmbH and Co AChE inhibitor2 Huperzine Mayo Foundation AChE inhibitor2 P-11012 Hoechst AG AChE inhibitor2 P-11149 Hoechst AG AChE inhibitor2 SR-46559A Sanofi Recherche SA M1 agonist2 T-588 Toyama Chemical Co Ltd ACh release stimulator2 TAK-147 Takeda Chemical Industries Ltd AChE inhibitor2 YM-796 Yamanouchi Pharmaceutical Co Ltd M1 agonist, nootropic agent2 Zifrosilone Hoechst Marion Roussel Inc AChE inhibitor2 Methanesulfonyl chloride University of Texas system AChE inhibitor2 S-9977 Servier AChE inhibitor1 LU-25109 H Lundbeck A/S mAChR agonist1 GTS-21 (DMXB-anabaseine) University of Florida Nicotinic agonist1 NS-2330 NeuroSearch A/S AChR agonist

ACh=acetylcholine; AChR=acetylcholine receptor; AChE=acetylcholinesterase; nAChR=nicotonic cholinergic receptor; M=muscarinic receptor.[118]

The cholinergic hypothesis of Alzheimers disease: a review 141

tive mechanisms.69 Thus, these compoundsreversibly bind to the hydrophobic region ofthe enzyme to allosterically modulate cata-lytic activity. Further, the inhibition producedby these compounds is mutually exclusive,suggesting that both compounds act at similarsites within the enzyme, although donepezil ismore potent and selective.70 71

This type of inhibition diVers from that pro-duced by the carbamates for example,rivastigmine and physostigmine derivativessuch as heptylphysostigmine. This class ofcompounds have been termed pseudoirre-versible ChE inhibitors, in that they are actu-ally cleaved by the enzyme, resulting in a cova-lent modification of the enzyme. Suchinhibition is non-competitive with ACh and isirreversible. However, the association of thecarbamate with the esteric site is transient(taking several minutes) due to both rapidmetabolism and the relative rapid rate ofdecarbamylation which regenerates ChE.7274 Afurther compound, metrifonate, inhibits ChEirreversibly. Metrifonate is a prodrug that isconverted into dichlorvos, an organophospho-rus ChE inhibitor with a very long duration ofinhibition (the half life is 52 days).75

The principal influence of the mechanism ofaction of enzyme inhibitors in the clinic relatesto their duration of action. A more theoreticalissue is the eVect of pronounced non-competitive inhibition on the rate of enzymesynthesis. Non-competitive inhibitors may pro-duce only slowly reversible ChE inhibition.The rate at which this inhibition is reversedmay be of the same order as the rate of enzymesynthesis.76 Thus, the long term eVects ofadministration of slowly reversible, or irrevers-ible, inhibitors on the overall cholinergic func-tion are diYcult to predict.

The selectivity of enzyme inhibition alsoplays a crucial part in determining thetherapeutic profile of any ChE inhibitor. In thisregard, several factors should be taken intoaccount. All compounds will possess a greateror lesser degree of selectivity, and many of thediVerences between compounds may be influ-enced by the actions of the compound otherthan its intended ChE inhibition. Not surpris-ingly, therapeutic agents developed as inhibi-tors of AChE, which is found primarily in neu-ral tissue, may also inhibitbutyrylcholinesterase (BuChE), which actsmainly in the periphery. Although the functionof BuChE remains unknown,77 clinical datawith selective and non-selective AChE inhibi-tors suggest the BuChE inhibition may beassociated with unwanted peripheral sideeVects,78 79 although to date, this remains anunproved empirical finding. However, com-pared with tacrine, less peripheral cholinergic-related side eVects have been found withdonepezil, as it is over 1000-fold more selectivefor AChE than BuChE.70 74 79 80 Thus, greaterbrain AChE inhibition may be achieved withdonepezil at the therapeutically eVective dosecompared with tacrine, increasing donepezilspotential clinical eYcacy.71

A further factor associated with the in vivopharmacology of mixed type ChE inhibitors is

that such compounds may interact with the siteat which ACh is captured within the AChEenzyme, and may also act at other sites thatbind or recognise Ach.81 82 Both tacrine anddonepezil displace the binding of selectiveligands from muscarinic and nicotinic AChreceptors,57 71 8385 although neither compoundhas significant activity at other neurotransmit-ter receptors. At muscarinic receptors, bothcompounds act as antagonists.71 However,these eVects only occur at concentrations of thecompounds significantly greater than thoseneeded to produce the required degree of ChEinhibition and are not therefore likely to haverelevance in the clinic.86

Donepezil, like tacrine, has been reported tohave eVects on other neurotransmitter systemsother than via receptors. For instance, donepe-zil is only 10-fold less potent than imipramineat inhibiting the uptake of serotonin.71 How-ever, unlike tacrine, some second generationChE inhibitors have been shown, using in vivomicrodialysis techniques to measure the extra-cellular concentration of neurotransmittersand their metabolites, to increase monoamineconcentrations in the cortex after administra-tion of therapeutic doses.87 88 These type ofeVects might be expected to influence aVectivestatesfor example, moodin a positive man-ner. Given that depression and aggression areimportant determinants of quality of life forpatients with Alzheimers disease and their car-ers, such eVects may have clinical relevance.1

CLINICAL TRIALS

First generation cholinesterase inhibitorsDuring the late 1980s and early 1990s, the firstcholinomimetic compound, tacrine, under-went a large number of clinical studies usingvarious doses and treatment periods rangingfrom a few days to 30 weeks. Tacrine was sub-sequently approved for use in some, but not all,countries. Evidence from three pivotal studiesof tacrine has established clearly the benefits ofChE treatment in patients with a diagnosis ofprobable Alzheimers disease.8991 Statisticallysignificant, dose related improvements onobjective performance based tests of cognition,clinician and caregiver rated global evaluationsof patient wellbeing, and also quality of lifemeasures have been reported.8992

Unfortunately, potentially serious adverseside eVects have limited the use of thiscompound. Both tacrine, and the carbamatephysostigmine, possess detrimental eVects onhepatic and cardiovascular function. Indeed,perhaps the most often documented reason forwithdrawal of tacrine is its potential hepatotox-icity. However, this eVect seems to be unrelatedto dose, and alanine aminotransferase (ALT)concentrations usually return to normal afterdrug withdrawal. In addition, many patientscan be successfully rechallenged with tacrineafter the enzymes return to normal.93

Among the other unwanted side eVects oftacrine, the most often occurring are thosecaused by overstimulation of the peripheralcholinergic system at or below 30% ChE inhi-bition reflecting its dose related tolerability.94

These side eVects are manifested predomi-

142 Francis, Palmer, Snape, et al

nantly by gastrointestinal tract discomfort andoveractivity, resulting in nausea, vomiting,abdominal pain, and diarrhoea. Doses oftacrine within the therapeutic range elicit suchside eVects in about 20% of tacrine treatedpatients.95 In one 30 week clinical trial, over50% of patients treated with tacrine discontin-ued treatment because of cholinergic relatedside eVects. In addition, over 70% of patientstitrated to the highest dosage of tacrine (160mg/day) failed to complete this 30 weekstudy.91 These incidental eVects limit the com-pounds maximum tolerated dose that may beadministered to patients, and therefore theextent of brain AChE inhibition that can beachieved.

Despite these limitations, a substantialnumber of patients, some 250 000300 000worldwide, have been exposed to tacrine. Con-sequently, although tacrine produces a mean-ingful benefit in a significant proportion ofpatients with Alzheimers disease, the questionhas been raised as to whether this approachrepresents a fair test of the cholinergic hypoth-esis. This issue has been considered in thedevelopment of second generation ChEinhibitors. Such ChE inhibitors have beendesigned to limit side eVect problems, and themaximum tolerated dose that can be achievedmay be determined more by the eVects of ChEinhibition itself.

Second generation cholinesterase inhibitorsAt least an equivalent level of benefit is likely tobe produced by the newer second generationChE inhibitors including donepezil,9698

rivastigmine,99 metrifonate,100102 galantamine103

and several other compounds. Such com-pounds show an eVect and magnitude ofbenefit of at least that reported for tacrine, butwith a more favourable clinical profile. Forexample, donepezil has a once daily dosageschedule and produces dose related significantimprovements in cognition and global func-tion, with over 80% of patients experiencing animprovement or no deterioration in cognition.Such responses should be viewed positively,considering the progressive, degenerative na-ture of the disease. In one 30 week randomised,double blind study of donepezil (5 or 10mg/day) versus placebo (n=150/group, 450total), statistically significant improvementswere obtained with both 5 and 10 mg/day ofdonepezil for the intent to treat analysis ofAlzheimers disease assessment scale(ADAS-cog104; p

hobbies and pastimes and, in some cases, evento go shopping and successfully return with therequired goods without having become lost.

Finally, evidence is emerging from clinicaltrials of cholinomimetic drugs that such drugsmay improve the abnormal non-cognitive,behavioural symptoms of Alzheimers disease.Thus, ChE inhibitors have been reported tosignificantly improve many manifestations ofbehavioural disturbance including agitation,apathy, hallucinations, and aberrant motorbehaviour101 109 and xanomeline, a selectivemuscarinic agonist, has been shown to improvevocal outbursts and psychotic symptoms.110

Further, such long term treatment with ChEinhibitors may delay or reduce the need fornursing home placement, and even reducemortality (a trend for reduced mortality hasbeen noted with tacrine).111 These findingsrequire confirmation, and it will be importantto establish whether this reduction in mortalityincreases the duration of the subsequent, moredependent stages of the disease. In addition,although nursing home placement may bedelayed, the eVect on the actual duration ofnursing home care should be ascertained.

CLINICAL USE

Choosing the right patientCholinomimetic treatment is targeted specifi-cally at patients with Alzheimers disease, albeitthat such treatment may be beneficial in otherdementias where a cholinergic deficit alsoexistsfor example, Lewy body disease. Trialsare currently underway to explore this possi-bility. The severity of the dementia is anotherimportant factor to be considered as currentlythese drugs have been assessed adequately inpatients with mild to moderately severeAlzheimers disease only, but again this is sub-ject to further evaluation and current practicemay change as clinical experience increases. Inaddition, it is essential to make a careful assess-ment of the patients illness to ensure that theyare likely to have Alzheimers disease. Primarycare physicians may screen for and recognisepatients with suspected Alzheimers diseasewithin the community, but often referral to aspecialist service is required. As there is nodefinitive diagnostic test for Alzheimers dis-ease, it is important to base a diagnosis ofprobable Alzheimers disease on careful con-sideration of the patients symptoms and signs,preferably using the Diagnostic and StatisticalManual of Mental Disorders, fourth edition(DSM IV)112 or National Institute of Neuro-logical and Communicative Disorders andStroke - Alzheimers Disease and Related Dis-orders Association work group (NINCDS-ADRDA) criteria,113 or an equivalent protocol.If properly applied, the accuracy rate ofdiagnosis using such criteria, confirmed atnecropsy, probably varies between 85% and95%,114 115 depending on the experience of thecentre in which the patient is assessed.

Decisions on continuing long term cholinesteraseinhibitor treatmentLess than half of the patients receiving ChEinhibitors achieve a clinically significant re-

sponse, although no further deterioration oreven a slowing of deterioration are desirableoutcomes, given the progressive, degenerativenature of the disease. Nevertheless, all patientswith Alzheimers disease should have theopportunity of a treatment trial of at least 3months in duration. Unfortunately, however, ithas not yet been possible to predict ordistinguish responders from non-responders.

In clinical trials, various psychometric out-come measures are used to assess the eYcacyof drug treatment for Alzheimers disease. Thecognitive subscale of the ADAS-cog and theCIBIC are used often to determine the efficacyof pharmacological agents. The ADAS-cogmeasures memory, orientation, attention, lan-guage, function, and praxis. CIBIC measurespatients global function in terms of general,cognitive, behavioural, and ADL domains (forexample, personal care and hobbies), and isdetermined by experienced clinicians, but mayincorporate input from the primary caregiverof the patient with Alzheimers disease (forexample, the CIBIC plus). A further assess-ment, the mini mental state examination(MMSE),116 is a short collection of cognitivetests that examines several areas of cognition. Itis widely used to measure the onset, progres-sion, and severity of Alzheimers disease in theclinical setting. The test is easy to administerand score, and can be used readily in a primarycare setting, both at the oYce and in thepatients home.

In clinical practice, it is rarely possible toundertake such a comprehensive assessment,but some degree of objectivity concerningtreatment eVect is, nevertheless, essential.Many physicians will rely on a simple generaltest such as the MMSE, coupled with a globalmeasure formed from the relative or other car-ers opinion about the response, and their ownassessment based on notes made at the firstassessment. Those patients that clearly benefitfrom treatment should continue. However,treatment should be terminated in thosepatients who show deterioration, albeit thatsuch patients must be closely monitored duringsuch periods due to reports of precipitousdecline after abrupt discontinuation. In addi-tion, drug free periods, between 36 monthsfor example, may be useful in evaluating theresponse to ChE inhibitors; deterioration dur-ing such periods would indicate that continuedtreatment is appropriate.117

ConclusionThe cholinergic hypothesis of Alzheimers dis-ease is based on the presynaptic deficits foundin the brains of patients with Alzheimersdisease and studies of the role of ACh in animaland human behaviour. Although it is now clearthat cholinergic dysfunction may not causecognitive impairment directly, but rather indi-rectly, by interfering with attentional process-ing, the hypothesis predicted that cholinomi-metic drugs would improve cognitive function.This prediction was not fully realised withcompounds such as physostigmine and tacrine,probably because the emergence of side eVectsthat may have constrained the dosing regimen

144 Francis, Palmer, Snape, et al

to sub-eYcacious doses. Poor tolerabilityseems to be less of an issue for the second gen-eration compounds of the type now beinglicensed for the treatment of Alzheimersdisease. With improved diagnosis, carefulpatient selection, and fewer side eVects, suchcompounds will establish if cholinomimetictherapy provides eVective and long lasting pal-liative therapy. Moreover, the emerging relationbetween neurotransmission and metabolism oftwo key proteins involved in Alzheimersdisease, APP and tau, raises the possibility thatsecond generation ChE inhibitors may alterdisease pathology and progression.

Work on this manuscript was supported by an educational grantfrom Eisai Inc and Pfizer Pharmaceuticals Group, Pfizer Inc. Wethank PPS International, Worthing, UK, for their assistance inthe development of this manuscript and Drs K Stanhope and MSheardown for helpful discussion.

1 Esiri MM. The basis for behavioural disturbances indementia. J Neurol Neurosurg Psychiatry 1996;61:12730.

2 Bowen DM, Smith CB, White P, et al. Neurotransmitter-related enzymes and indices of hypoxia in senile dementiaand other abiotrophies. Brain 1976;99:45996.

3 Davies P, Maloney AJF. Selective loss of central cholinergicneurones in Alzheimers disease. Lancet 1976;ii:1403.

4 Perry EK, Gibson PH, Blessed G, et al. Neurotransmitterenzyme abnormalities in senile dementia. Choline acetyl-transferase and glutamic acid decarboxlyase activities innecropsy brain tissue. J Neurol Sci 1977;34:24765.

5 Rylett RJ, Ball MJ, Colhuon EH. Evidence for high aYnitycholine transport in synaptosomes prepared from hippoc-ampus and neocortex of patients with Alzheimers disease.Brain Res 1983;289:16975.

6 Nilsson L, Nordberg A, Hardy JA, et al. Physostigminerestores [3H]-acetylcholine eZux from Alzheimer brainslices to normal level. J Neural Transm 1986;67:27585.

7 Whitehouse PJ, Price DL, Struble RG, et al. Alzheimersdisease and senile dementia: loss of neurones in basal fore-brain. Science 1982;215:12379.

8 Drachman DA, Leavitt J. Human memory and the choliner-gic system. Arch Neurol 1974;30:11321.

9 Bartus RT, Dean RL, Beer B, et al. The cholinergic hypoth-esis of geriatric memory dysfunction. Science 1982;217:40817.

10 Mann DMA. Pyramidal nerve cell loss in Alzheimersdisease. Neurodegeneration 1996;5:4237.

11 DeKosky ST, ScheV SW, Styren SD. Structural correlates ofcognition in dementia: quantification and assessment ofsynapse change. Neurodegeneration 1996;5:41721.

12 Francis PT, Sims NR, Procter AW, et al. Cortical pyramidalneurone loss may cause glutamatergic hypoactivity andcognitive impairment in Alzheimers disease: investigativeand therapeutic perspectives. J Neurochem 1993;60:1589604.

13 Perry EK, Tomlinson BE, Blessed G, et al. Correlation ofcholinergic abnormalities with senile plaques and mentaltest scores in senile dementia. BMJ 1978;2:14579.

14 Wilcock GK, Esiri MM, Bowen DM, et al. Alzheimers dis-ease. Correlation of cortical choline acetyltransferase activ-ity with the severity of dementia and histological abnor-malities. J Neurol Sci 1982;57:40717.

15 Sims NR, Bowen DM, Allen SJ, et al. Presynaptic choliner-gic dysfunction in patients with dementia. J Neurochem1983;40:5039.

16 Rossor MN, Garrett NJ, Johnson AL, et al. A postmortemstudy of the cholinergic and GABA systems in seniledementia. Brain 1982;105:31330.

17 Lowe SL, Bowen DM, Francis PT, et al. Antemortemcerebral amino acid concentrations indicate selectivedegeneration of glutamate-enriched neurons in Alzheimersdisease. Neuroscience 1990;38:5717.

18 Rossor M. Iversen LL. Non-cholinergic neurotransmitterabnormalities in Alzheimers disease. Br Med Bull 1986;42:7044.

19 Francis PT, Bowen DM, Lowe SL, et al. Somatostatin con-tent and release measured in cerebral biopsies fromdemented patients. J Neurol Sci 1987;78:116.

20 Palmer AM, Stratmann GC, Procter AW, et al. Possible neu-rotransmitter basis of behavioural changes in Alzheimersdisease. Ann Neurol 1988;23:61620.

21 Chen CPL, Alder JT, Bowen DM, et al. Presynapticserotonergic markers in community-acquired cases ofAlzheimers disease: correlation with depression andneuroleptic medication. J Neurochem 1996;66:15928.

22 Bowen DM, Benton JS, Spillane JA, et al. Cholineacetyltransferase activity and histopathology of frontal neo-cortex from biopsies of demented patients. J Neurol Sci1982;57:191202.

23 Neary D, Snowden JS, Mann DM, et al. Alzheimers disease:a correlative study. J Neurol Neurosurg Psychiatry 1986;49:22937.

24 Palmer AM, Procter AW, Stratmann GC, et al. Excitatoryamino acid-releasing and cholinergic neurones inAlzheimers disease. Neurosci Lett 1986;66:199204.

25 Kish SJ, Robitaille Y, El-Awar M, et al. Non-Alzheimer-typepattern of brain choline acetyltransferase reduction indominantly inherited olivopontinecerebellar atrophy. AnnNeurol 1989;26:3627.

26 Whitehouse PJ, Martion AM, Marcus KA, et al. Reductionsin acetylcholine and nicotine binding in several degenera-tive diseases. Arch Neurol 1988;45:7224.

27 Nordberg A, AlafuzoV I, Winblad B. Nicotinic andmuscarinic subtypes in the human brain: changes withaging and dementia. J Neurosci Res 1992;31:10311.

28 Warpman U, AlafuzoV I, Nordberg A. Coupling ofmuscarinic receptors to GTP proteins in postmortemhuman brainalterations in Alzheimers disease. NeurosciLett 1993;150:3943.

29 Chessell IP, Francis PT, Bowen DM. Changes in corticalnicotinic acetylcholine receptor numbers following unilat-eral destruction of pyramidal neurones by intrastriatalvolkensin injection. Neurodegeneration 1995;4:41524.

30 McCormick DA, Prince DA. Two types of muscarinicresponses to acetycholine in mammalian cortical neurones.Proc Natl Acad Sci USA 1985;82:63448.

31 Halliwell JV. M-current in human neocortical neurones.Neurosci Lett 1986;67:16.

32 Dijk SN, Francis PT, Stratmann GC, et al. Cholinomimeticsincrease glutamate outflow by an action on the corticostri-atal pathway: implications for Alzheimers disease. J Neuro-chem 1995;65:21659.

33 Hyman BT, Van Hoesen GW, Damasio AR. Alzheimersdisease: glutamate depletion in the hippocampal perforantpathway zone. Ann Neurol 1987;22:3740.

34 Procter AW, Palmer AM, Francis PT, et al. Evidence ofglutamatergic denervation and possible abnormal metabo-lism in Alzheimers disease. J Neurochem 1988;50:790802.

35 Procter AW, Francis PT, Holmes C, et al. APP isoformsshow correlations with neurones but not with glia in brainsof demented subjects. Acta Neuropathol 1994;88:54552.

36 Procter AW. Neurochemical correlates of dementia. Neuro-degeneration 1996;5:4037.

37 Brun A. An overview of light and electron microscopicchanges in Alzheimers disease, the standard reference.New York: Macmillan, 1983:3747.

38 Lewis DA, Campbell MJ, Terry RD, et al. Laminar andregional distributions of neurofibrillary tangles and neuriticplaques in Alzheimers disease a quantitative study of visualand auditory cortices. J Neurosci 1987;7:1799809.

39 Rapoport SI, Hatanpaa K, Brady DR, et al. Brain energymetabolism, cognitive function and down-regulated oxida-tive phosphorylation in Alzheimer disease. Neurodegenera-tion 1996;5:4736.

40 DeKosky ST, ScheV SW. Synapse loss in frontal cortexbiopsies in Alzheimers disease: correlation with cognitiveseverity. Ann Neurol 1990;27:45764.

41 Terry RD, Masliah D, Salman N, et al. Psychometrics andtheir correlation with physical aspects of the Alzheimersbrain. Neurobiol Aging 1990;11:266.

42 Wilcock GK, Esiri MM. Plaques, tangles and dementia. JNeurol Sci 1982;56:34356.

43 Najlerahim A, Bowen DM. Biochemical measurements inAlzheimers disease reveal a necessity for improvedneuroimaging techniques to study metabolism. Biochem J1988;251:3058.

44 Gustafson L, Edvinsson L, Dahlgren N, et al. Intravenousphysostigmine treatment of Alzheimers disease evaluatedby psychometric testing, regional cerebral blood flow(rCBF) measurement, and EEG. Psychopharmacology1987;93:315.

45 Geaney DP, Soper N, Shepstone BJ, et al. EVect of centralcholinergic stimulation on regional cerebral blood flow inAlzheimers disease. Lancet 1990;335:14847.

46 Hardy J, Allsop D. Amyloid deposition as the central eventin the aetiology of Alzheimers disease. Trends Pharmacol Sci1991;12:3838 .

47 Kar S, Issa AM, Seto D, et al. Amyloid -peptide inhibitshigh-aYnity choline uptake and acetylcholine release in rathippocampal slices. J Neurochem 1998;70:217987.

48 Auld DS, Kar S, Quirion R. -Amyloid peptides as directcholinergic neuromodulators: a missing link? Trends Neuro-sci 1998;21:439.

49 Nitsch RM. From acetylcholine to amyloid: neurotransmit-ters and the pathology of Alzheimers disease. Neurodegen-eration 1996;5:47782.

50 Emmerling MR, Schwarz RD, Spiegel K, et al. Newperspectives on developing muscarinic agonists for treatingAlzheimers disease. ID Research Alert 1997;2:18794.

51 Clarke NA, Webster M, Francis PT, et al. -Amyloidprecursor protein-like immunoreactivity can be altered inhumans by drugs aVecting neurotransmitter function. Neu-rodegeneration 1993;2:2438.

52 Lovestone S, Reynolds H, Latimer D, et al. Alzheimersdisease-like phosphorylation of the microtubule-associatedprotein tau by glycogen synthase kinase-3 in transfectedmammalian cells. Curr Biol 1994;4:107786.

53 Sadot E, Gurwitz D, Barg J, et al. Activation of m1muscarinic acetylcholine receptor regulates tau phosphor-ylation in transfected PC12 cells. J Neurochem 1996;66:87780.

54 Palmer AM, Gershon S. Is the neurochemical basis ofAlzheimers disease cholinergic or glutamatergic? FASEB1990;4:274552.

The cholinergic hypothesis of Alzheimers disease: a review 145

55 Andrews JS, Grutzner M, Stephens DN. EVects of choliner-gic and non-cholinergic drugs on visual discrimination anddelayed visual discrimination performance in rats. Psychop-harmacology 1992;106:52330.

56 Poorheidari G, Stanhope KJ, Pratt JA. EVects ofthe potassium channel blockers, apamin and4-aminopyridine, on scopolamine-induced deficits in thedelayed matching to position task in rats: a comparisonwith the cholinesterase inhibitor E2020. Psychopharmacol-ogy 1998;135:24255.

57 Stanhope KJ, McLenachan A, Dourish CT. Dissociationbetween cognitive and motor/motivation deficits in thedelayed matching to position test: eVects of scopolamine,8-OH-DPAT and EAA antagonists. Psychopharmacology1995;122:26880.

58 Dunne MP, Hartley LR. Scopolamine and the control ofattention in humans. Psychopharmacology 1986;89:947.

59 Warburton DM, Rusted JM. Cholinergic control ofcognitive resources. Neuropsychobiology 1993;28:436.

60 Hagan JJ, Salamone JD, Simpson J, et al. Place navigation inrats is impaired by lesions of medial septum and diagonalband but not nucleus basalis magnocellularis. Behav BrainRes 1988;27:920.

61 Van der Staay FJ, Raaijmakers WGM, Lammers AJJC, et al.Selective fimbra lesions impair acquisition of working andreference memory of rats in a complex spatial discrimina-tion task. Behav Brain Res 1989;32:15161.

62 Dunnett SB, Everitt BJ, Robbins TW. The basal forebrain-cortical cholinergic system: interpreting the functionalconsequences of excitotoxic lesions. Trends Neurosci 1991;14:494501.

63 Dawson GR, Heyes CM, Iversen SD. Pharmacologicalmechanisms and animal models of cognition. Behav Phara-macol 1992;3:28597.

64 Muir JL, Everitt BJ, Robbins TW. AMPA-induced excito-toxic lesions of the basal forebraina significant role forthe cortical cholinergic system in attentional function. JNeurosci 1994;14:231326.

65 Muir JL, Everitt BJ, Robbins TW. Reversal of visualattentional dysfunction following lesions of the cholinergicbasal forebrain by physostigmine and nicotine but not bythe 5-HT3 receptor antagonist, ondansetron. Psychophar-macology 1995;118:8292.

66 Sahakian BJ, Coall JT. Nicotine and tetra-hydroaminoacridine: evidence for improved attention inpatients with dementia of the Alzheimer type. Drug Dev Res1994;31:808.

67 Treanor J, Beck K, Hefti F. Neuronal growth factors andAlzheimers disease. In: Dawbarn D, Allen SJ, eds. Neurobi-ology of Alzheimers disease. Bristol: Bios Scientific Publish-ers, 1995:24959.

68 Baker HF, Ridley RM. Cholinergic grafts in primates: therole of ACh in cognition. In: Dawbarn D, Allen SJ, eds.Neurobiology of Alzheimers disease. Bristol: Bios ScientificPublishers, 1995:22342.

69 Nochi S, Asakawa N, Sato T. Kinetic study on the inhibitionof acetylcholinesterase by 1-benzyl-4-[(5,6-dimethoxy-l-indanon)-2-yl] methylpiperidine hydrochloride (E2020).Biol Pharm Bull 1995;18:11457.

70 Rogers SL, Yamanishi Y, Yamatsu K. The pharmacology ofa piperidine cholinesterase inhibitor. In: Becker R,Giacobini E, eds. Cholinergic basis for Alzheimer therapy.Boston: Birkhuser, 1991:31420.

71 Snape MF, Anderson SMP, Misra A, et al. A comparison ofthe cholinesterase-inhibitors tacrine and E-2020. Br JPharmacol 1996;117:P163.

72 Dawson RM. Reversibility of the inhibition of acetylcho-linesterase by tacrine. Neurosci Lett 1990;118:857.

73 Moriearty PL, Becker RE. Inhibition of human brain andRBC acetylcholinesterase (AChE) by heptylphysostigmine(HPTL). Methods Find Exp Clin Pharmacol 1992;14:61521.

74 Sherman KA. Pharmacodynamics of oral E2020 and tacrinein humans: novel approaches. In: Becker R, Giacobini E,eds. Cholinergic basis for Alzheimer therapy. Boston:Birkhuser, 1991:3218.

75 Doraiswamy PM. Current cholinergic therapy for symp-toms of Alzheimers disease. Primary Psychiatry 1996;3:5668.

76 Homor G, Kasa P. Acetylcholinesterase resynthesis afterDFP poisoning: histochemical and biochemical study. ActaHistochem 1978;62:293301.

77 Taylor P. The cholinesterases. J Biol Chem 1991;266:40258.

78 Thomsen T, Zendeh B, Fischer JP, et al. In vitro eVects ofvarious cholinesterase inhibitors on acetyl- and butyrylcho-linesterase of healthy volunteers. Biochem Pharmacol 1991;41:13941.

79 Sugimoto H, Iimura Y, Yamanishi Y, et al. Synthesis andstructure-activity relationships of acetylcholinesterase in-hibitors 1-benzyl-4-[(5,6-dimethoxy-1-oxoindan-2-yl)methyl]piperidine hydrochloride and related compounds. JMed Chem 1995;38:48219.

80 Hunter AJ, Murray TK, Jones JA, et al. The cholinergicpharmacology of tetrahydroaminoacridine in vivo and invitro. Br J Pharmacol 1989;98:7986.

81 Sussman JL, Harel M, Silman I. 3-Dimensional structureof acetylcholinesterase and of its complexes withanticholinesterase drugs. Chem Biol Interact 1993;87:18797.

82 Harel M, Schalk I, Ehretsabatier L, et al. Quaternary ligand-binding to aromatic residues in the active-site gorge of ace-tylcholinesterase. Proc Natl Acad Sci USA 1993;90:90315.

83 Clarke PBS, Reuben M, Elbizri H. Blockade of nicotinicresponses by physostigmine, tacrine and othercholinesterase-inhibitors in rat striatum. Br J Pharmacol1994;111:695702.

84 Nilsson L, Adem A, Hardy J, et al. Do tetrahydroaminoacri-dine (THA) and physostigmine restore acetylcholine-release in Alzheimer brains via nicotinic receptors? J NeuralTransm Suppl 1987;70:35768.

85 Perry EK, Smith CJ, Court JA, et al. Interaction of 9-amino-1,2,3,4-tetrahydroamino-acridine (THA) with human cor-tical nicotinic and muscarinic receptor-binding in vitro.Neurosci Lett 1988;91:2116.

86 Pearce BD, Potter LT. EVects of tetrahydroaminoacridineon M1 and M2 muscarine receptors. Neurosci Lett 1988;88:2815.

87 Zhu XD, Cuadra G, Brufani M, et al. EVects of MF-268, anew cholinesterase inhibitor, on acetylcholine andbiogenic-amines in rat cortex. J Neurosci Res 1996;43:1206.

88 Baldwin HA, Desouza RJ, Sarna GS, et al. Measurementsof tacrine and monoamines in brain by in vivomicrodialysis argue against release of monoamines bytacrine at therapeutic doses. Br J Pharmacol 1991;103:194650.

89 Davis KL, Thal LJ, Gamzu ER, et al. A double-blind,placebo-controlled multicenter study of tacrine forAlzheimers disease. N Engl J Med 1992;327:12539.

90 Farlow M, Gracon SI, Hershey LA, et al. A controlled trialof tacrine in Alzheimers disease. The Tacrine StudyGroup. JAMA 1992;268:25239.

91 Knapp JM, Knopman DS, Soloman PR, et al. A 30-weekrandomized controlled trial of high-dose tacrine in patientswith Alzheimers disease. The Tacrine Study Group. JAMA1994;271:98591.

92 Eagger SA, Harvey RJ. Clinical heterogeneity: responders tocholinergic therapy. Alzheimer Dis Assoc Disord 1995;9(suppl 2):3742.

93 Wilcock GK, Scott M, Pearsall T. Long-term use of tacrine[letter]. Lancet 1994;11:56771.

94 Becker RE, Moriearty P, Unni L. The second generationof cholinesterase inhibitors: clinical and pharmacologicaleVects. In: Becker RE, Giacobini E, eds. Cholinergicbasis for Alzheimer therapy. Boston: Birkhuser, 1991:26396.

95 Schneider LS. Clinical pharmacology of aminoacridines inAlzheimers disease. Neurology 1993;43(suppl 14):S6479.

96 Rogers SL, FriedhoV ST, and the Donepezil Study Group.The eYcacy and safety of donepezil in patients withAlzheimers disease: results of a US multicenter, rand-omized, double-blind, placebo-controlled trial. Dementia1996;7:293303.

97 Rogers SL, Farlow MR, Mohs RC, et al, and the DonepezilStudy Group. A 24-week, double-blind, placebo-controlledtrial of donepezil in patients with Alzheimers disease. Neu-rology 1998;50:13645.

98 Rogers SL, Doody RS, Mohs RC, et al, and the DonepezilStudy Group. Donepezil improves cognition and globalfunction in Alzheimers disease: a 15-week, double-blind,placebo-controlled study. Arch Int Med 1998;158:102131.

99 Corey-Bloom J, Anand R, Veach J, for the ENA 713 B352Study Group. A randomized trial evaluating the eYcacyand safety of ENA 713 (rivastigmine tartrate), a newacetylcholinesterse inhibitor, in patients with mild to mod-erately severe Alzheimers disease. Int J Geriatr Psychophar-macol 1998;1:5565.

100 Cummings L, Cyrus PA, Bieber F, et al. Metrifonate treat-ment of the cognitive deficits of Alzheimers disease.Neurology 1998;50:121421.

101 Morris JC, Cyrus PA, Orazem J, et al. Metrifonate benefitscognitive, behavioral, and global function in patients withAlzheimers disease. Neurology 1998;50:122230.

102 Knopman DS. Metrifonate for Alzheimers disease. Is thenext cholinesterase inhibitor better? Neurology 1998;50:12035.

103 Fulton B, Benfield P. Galanthamine. Drugs Aging 1996;9:605.

104 Rosen WG, Mohs RC, Davis K. A new rating scale forAlzheimers disease. Am J Psychiatry 1984;141:135664.

105 Schneider LS, Olin JT, Doody RS, et al. Validity andreliability of the Alzheimers Disease Cooperative Study-Clinical Global Impression of Change. Alzheimer Dis AssocDisord 1997;11(suppl 2): S2232.

106 FriedhoV LT, Rogers SL. Donepezil lengthens time to lossof activities of daily living in patients with mild to moderateAlzheimers diseaseresults of a preliminary evaluation.Neurology 1997;48:A100(P02.026).

107 Rogers SL, FriedhoV LT. Long-term eYcacy and safety ofdonepezil in the treatment of Alzheimers disease: aninterim analysis of the results of a US multicentre openlabel extension study. Eur Neuropsychopharmacol 1998;8:6775.

108 Mihara M, Ohnishi A, Tomono Y, et al. Pharmacokineticsof E2020, a new compound for Alzheimers disease, inhealthy male volunteers. Int J Clin Pharmacol Ther Toxicol1993;31:2239.

109 Kaufer DI, Catt K, Pollock BG, et al. Donepezil inAlzheimers disease: relative cognitive and neuropsychiatricresponses and impact on caregiver distress. Neurology1998;50(suppl 4):A89.

110 Bodick NC, OVen WW, Levey AI, et al. EVects of xanome-line, a selective muscarinic receptor agonist, on cognitivefunction and behavioral symptoms in Alzheimer disease.Arch Neurol 1997;54:46573.

146 Francis, Palmer, Snape, et al

111 Knopman D, Schneider L, Davis K, et al. Long-termtacrine (Cognex) treatment - eVects on nursing-homeplacement and mortality. Neurology 1996;47:16677.

112 American Psychiatric Association. Diagnostic and statisticalmanual of mental disorders. 4th ed. Washington, DC: APA,1994.

113 McKhann G, Drachman D, Folstein M, et al. Clinicaldiagnosis of Alzheimers disease: report of the NINCDS-ADRDA work group. Neurology 1984;34:93944.

114 Galasko D, Hansen LA, Katzman R, et al. Clinical-neuropathological correlations in Alzheimers disease andrelated dementias. Arch Neurol 1994;51:88895.

115 Gearing M, Mirra SS, Hedreen JC, et al. The Consortiumto Establish a Registry for Alzheimers Disease (CERAD).Part X. Neuropathology confirmation of the clinicaldiagnosis of Alzheimers disease. Neurology 1995;45:4616.

116 Folstein NF, Folstein SE, McHugh PR. Mini mental state:a practical method for grading the cognitive state ofpatients for the clinician. J Psychiatr Res 1975;12:18998.

117 Lovestone S, Graham N, Howard H. Guidelines on drugtreatments for Alzheimers disease. Lancet 1997;350:2323.

118 Anonymous. Alzheimers diseasedrug status update.Current Research in Alzheimers Disease 1998;3:26571.

The cholinergic hypothesis of Alzheimers disease: a review 147