AUGMENTING CHOLINERGIC FUNCTION
Since ACh appears to be important in memory processing and as its concentration is significantly reduced in appropriate brain areas in AzD then augmenting its action should at least improve memory function. ACh activity may be increased by
(1) Enhancing its synthesis (and presumed release) through giving the precursor choline
(2) Stopping its degradation by cholinesterase with anticholinesterase drugs
(3) Reproducing its action with appropriate agonists — (a) muscarinic, (b) nicotinic
Approaches (1) and (2) depend on sufficient cholinergic function remaining to make its supplementation feasible, while all three methods suffer from the fact that not only does ACh have other central effects (e.g. in striatum), but also numerous peripheral para-sympathomimetic ones, such as increased bronchial and gastric secretion or reduced heart rate.
This requires the provision of the precursor choline which is often given as lecithin (phosphatidyl choline), a natural source of choline found in many foods such as eggs and fish. Large doses (9-10 g) have to be given, probably to overcome the body's natural ability to restrict plasma choline levels, and the fact that only a very small percentage is converted to ACh. Brain penetration is modest but uptake into cholinergic nerve terminals is through a sodium-dependent high-affinity system that is normally adequately supplied and possibly saturated with choline. In any case, ACh can only be synthesised from choline in cholinergic nerve terminals, many of which will have degenerated, and just increasing the activity of those remaining may not be adequate. Whether choline could reverse the choline leakage and resulting autocannibalism (see above) of cholinergic neurons is not known.
ACh is metabolised extraneuronally by the enzyme acetylcholinesterase, to reform precursor choline and acetate. Blocking its activity with various anticholinesterases has been widely investigated and some improvement in memory noted. Such studies have invariably used reversible inhibition because of the toxicity associated with long-term irreversible inhibition of the enzyme. Physostigmine was the pilot drug. It is known to improve memory in animals and some small effects have been seen in humans (reduces number of mistakes in word-recall tests rather than number of words recalled), but it really needs to be given intravenously and has a very short half-life (30min).
The limited effectiveness of physostigmine did, however, encourage the development of longer-acting orally effective anticholinesterases such as tacrine (tetrahydroamino-acrydine), velnacrine and donepezil.
Clinical evaluation of anticholinesterases and other drugs in AzD
The newer anticholinesterases have all been subject to large and often multicentred trials. These take various forms but generally include an initial assessment of disease severity over a few weeks while on placebo alone, then a drug-dose evaluation before the chosen drug dose(s) is compared directly with placebo for some weeks in two groups. Confirmation of any drug effect is usually obtained by finishing with all patients on placebo. Although performed double-blind generally, only patients that respond in the early evaluation period enter the final drug trial and those with severe AzD are excluded altogether. Results from a simpler Phase III drug study showing some efficacy for donepezil are shown in Fig. 18.4 (Rogers et al. 1998).
The evaluation of drug effectiveness in AzD is not without its difficulties. There is a need to record changes in both cognitive function and general performance. Two primary measures are the Alzheimer's Disease Assessment Scale of cognition (ADAS-cog) and the Clinician's Global Impression of Change (CGIC). The former measures such things as memory, language, orientation, reason and praxis, on a 0-70 scale range. The higher the score, the more severe the condition, and as most patients normally decline at the acquisition rate of 5-10 extra points a year, any reduction of 4 or more points is considered a drug effect. The CGIC scale, as its name implies, is a more global measure of patient function not only in cognition but also in general behaviour and daily living obtained by the clinician interviewing both patients and carers. On a 7-point scale, improvement is represented by 1, worsening by 7 and no change by 4. Other measures include the patient's own evaluation of quality of life (QoL) noting their general feelings and ability to eat, sleep and relate to others. Generally improvements in
Figure 18.4 Clinical assessment of the efficacy of the anticholinesterase drug, donepezil, in Alzheimer's disease. Summary redrawing of some of the results of a large double-blind placebo-controlled trial by Rogers et al. (1998) © Lippincott Williams & Wilkins in which the effect of donepezil (10mgday~') was tested on men and women over 50 years with uncomplicated Alzheimer's disease (N = 157) compared with placebo (n = 162), using a number of measures including the Alzheimer's Disease Association Assessment Scale of cognition (ADAS-cog) and a more global assessment (CIBIC plus) equivalent to the Clinician's Interview-based Impression of Change scale based on clinical, patient and family assessment of cognition and behaviour. The results show that donezepil has had a significant (p = 0.009-0.0001) beneficial effect by both the ADAS (•—•) and CIBIC (■-■) assessments, when compared with placebo (O-O and □-□) from 12th to 24th week. Two further features, characteristic of such therapy, are also apparent; (i) the drug has a greater effect on cognition (ADAS) than on overall state of health (CIBIC) and (ii) it does not retard the progress of the disorder (no difference between drug and placebo groups 6 weeks after cessation of drug). ABAS scores range from 0 to 70 (min-max symptoms) with patients normally deteriorating at a rate of 7-11 extra points per year so that any reduction in that rate constitutes an improvement. The CIBIC scale scores 1-7, with 1 = marked improvement, 4 = no change and 7 = worsening cognition (ADAS-cog scores) are more easily achieved than in the overall quality of life (CGIC) (see Fig. 18.4) and is a useful reminder that AzD is not just a loss of memory.
The initial enthusiasm for tacrine and velnacrine, which are the anticholinesterases most studied clinically, has been tempered by the fact that not all patients respond. Most show the peripheral parasympathomimetic effects of cholinesterase inhibition, e.g. dyspepsia and diarrhoea, as well as nausea and vomiting, and about half of the patients develop hepatotoxicity with elevated levels of plasma alanine transaminase. While some peripheral effects can be attenuated with antimuscarinics that do not enter the brain, these add further side-effects and the drop-out rate from such trials is high (<75%) in most long-term studies. Donepezil appears to show less hepatotoxicity but its long-term value remains to be determined.
Generally, anticholinesterases produce some improvement in cognitive function and the quality of life in the early stages of AzD but that needs to be balanced against side-effects.
Some of the cognitive improvements with tacrine, which is chemically related to amidopyridine, may be due to blockage of K+ channels.
Use of agonists
Since most postynaptic cholinergic receptors in the brain are muscarinic and as they do not appear to be reduced in AzD, despite some degeneration of pyramidal neurons, the use of muscarinic agonists could be worth while. Early studies with bethanecol, arecoline and oxotremorine, mixed M1 and M2 agonists, showed little benefit and newer drugs have not been much better. Peripheral cholinergic effects are a problem and central infusion, which has been tried with bethanecol to no great effect, is hardly a practical proposition. There is, however, a realisation that more appropriate drug or drug combinations could be developed and tried. Thus, theoretically anyway, the requirement is for a specific M1 agonist that readily crosses the blood-brain barrier. Avoiding M2 receptor stimulation will also mean no activation of presynaptic auto-receptors and counterproductive inhibition of ACh release, and fewer peripheral effects. These latter could also be avoided with an M1 antagonist that does not cross the blood-brain barrier. Even then successful therapy may be negated by a requirement for ACh to be released physiologically from appropriate neurons.
Despite the paucity of nicotinic receptors in the brain there is considerable evidence that AzD is less common among smokers. Whether this is due to the action of inhaled nicotine is uncertain, but nicotine is known to stimulate presynaptic receptors on cholinergic nerve terminals which, unlike the muscarinic ones, result in increased ACh release.
If long-term potentiation (LTP) is important in memory function and as it can be blocked by glutamate NMDA antagonists (see above) then increasing NMDA activity is of putative value in AzD. In reality this presents a problem because overstimulation of the receptor could not only increase neuronal function up to convulsive level but even cause neurotoxicity. Briefly, NMDA applied to rat cortex causes retrograde degeneration of cholinergic neurons in nucleus basalis while NMDA antagonists prevent anoxic destruction of cultured hippocampal neurons and brain damage caused by cerebral vascular occlusion in rodents. The ischaemia the latter produces causes such an excessive neuronal discharge and release of glutamate that the intense activation of NMDA receptors produces a prolonged neuronal depolarisation, Ca2+ entry and cell death. Possibly a weak partial NMDA agonist, or a drug acting at one of the NMDA receptor subsites (see Chapter 10) like that for glycine, may be of some value. Whether glutamate-induced neuronal death, which is enhanced by ^-amyloid, plays any part in the actiology of AzD is uncertain but controlling glutamate activity so that it can be increased enough to facilitate memory processes without undue excitation of neurons will be difficult.
Although there is no neuropathological evidence to implicate GABA in AzD it is known that agonists at the benzodiazepine receptor site not only augment GABA function but also cause amnesia. So it is possible that an inverse agonist, or perhaps even an antagonist, for the benzodiazepine receptor could have the opposite effect and improve memory. In humans, one antagonist, the ¿ carboline derivative ZK93426, showed some improvement in learning and memory tests. It also improves acquisition in animal-learning tests and counteracts the impairment caused by scopolamine, as does the ¿-carboline inverse agonist DMCM. The fear of inducing anxiety or even convulsions with inverse benzodiazepine agonists has prompted the evaluation of partial inverse agonists (see Abe, Takeyama and Yoshimura 1998).
There have been few attempts to manipulate the monoamines in AzD and those using selegiline, the MOAB inhibitor, have shown little effect although the 5-HT3 antagonist, ondansetron, may give a slight improvement.
Despite the clear loss of somatostatin in AzD a synthetic analogue L-363586 had no beneficial effect on memory loss.
Was this article helpful?