Acetylcholinesterase Inhibitors and Alzheimer’s Disease

Acetylcholinesterase Inhibitors and Alzheimer’s Disease 9

Acetylcholinesterase Inhibitors and Alzheimer’s Disease H Kaduszkiewicz and H van den Bussche, University Medical Center Hamburg-Eppendorf, Hamburg, Germany ã 2009 Elsevier Ltd. All rights reserved.

Introduction The age specific prevalence of Alzheimer’s disease (AD) doubles every 5 years of life, affecting 1.2% of those aged 65–69, 2.8% of those aged 70–74, 6.0% of those aged 75–79, etc. The disease causes irreversible memory loss, behavioral and cognitive decline, personality changes, and a decreasing ability to cope with everyday life. Amyloid plaques, neurofibrillary tangles, and a loss of neurons are the typical cellular pathology observed in the central nervous system of AD patients, though these changes are not pathognomonic. In the 1970s, postmortem examinations of the brains of patients with largely end-stage AD also reported markedly reduced activity of choline acetyltransferase, an acetylcholine-synthesizing enzyme. The cholinergic deficits were reported to correlate with cognitive deficits and disease severity. They were also linked to neuropsychiatric symptoms of AD, when medial frontal and limbic cholinergic deficits were identified. The discovery of widespread loss of cholinergic innervation triggered development of cholinesterase inhibitors, which aim to raise acetylcholine levels in the brain by blocking the enzymes metabolizing this molecule. Donepezil, rivastigmine, and galantamine are the three acetylcholinesterase (AChE) inhibitors licensed for the treatment of mild to moderately severe AD. Their characteristics are listed in Table 1.

Preclinical Testing Originally, preclinical testing of the AChE inhibitors was promising. For example, donepezil at doses of 2.5 mg kg 1 showed maximum increases of the extracellular acetylcholine concentration in the hippocampus of rats of 499% of the prelevel at about 1.5 h after administration. The time courses of brain AChE inhibition with donepezil correlated strongly with the extracellular acetylcholine-increasing action. Donepezil demonstrated efficacy in tests of reference memory in animals, but had less consistent activity in tests of working memory. AChE activity was also inhibited in samples of postmortem human brain, fresh brain cortex biopsies, and human erythrocytes. In addition, rapid, sustained, and dose-dependent inhibition of cerebral spinal fluid AChE was demonstrated in AD patients.

Some findings suggested a neuroprotective potential of AChE inhibitors. For example, donepezil was found to attenuate A-beta(25-35)-induced toxicity in rat pheochromocytoma PC12 cells in clinically relevant concentrations. Additional hope came from findings of special action of galantamine directly interacting with nicotinic acetylcholine receptors. These allosterically potentiating ligands sensitize nicotinic receptors by increasing the probability of channel opening induced by acetylcholine and nicotinic agonists and by slowing down receptor desensitization.

Clinical Efficacy Despite the promising preclinical findings, the evaluation of clinical efficacy did not reveal striking effects. In fact, the study results are interpreted so diversely that 10 years after licensure, the clinical efficacy of AChE inhibitors is still controversial. In this article, the manifold discussions on the study results on AChE inhibitors will be presented, as they are crucial for the estimation of clinical efficacy. There is no direct way to deduce a clinical benefit from study results. In fact, there is still a long way to go to accomplish this, namely: 1. Selection of relevant studies 2. Identification of the methodological quality of the studies 3. Assessment of the relevance of the methodological quality for the results 4. Discussion of the external validity of the trials 5. Determination of the clinical relevance of effects and side effects measured in the trials 6. Comprehension of the clinical experience of patients, caregivers, and health personnel Not until these steps are taken is an estimation of clinical benefit possible. Selection of Relevant Studies

In order to assess the efficacy of cholinesterase inhibitors, there is wide consensus that the analysis of randomized controlled double blind trials is the first choice. In these studies the agent should have been tested against placebo, and clinical parameters should have been used as outcome measures. To date, 18 randomized controlled trials (RCTs) on donepezil, five on rivastigmine, and seven on galantamine have been published. The majority of these studies included patients with mild to moderate stages of AD; the minority focused on patients in the severe stage.

10 Acetylcholinesterase Inhibitors and Alzheimer’s Disease Table 1 Characteristics of the three licensed AChE inhibitors for the treatment of AD patients Donepezil hydrochloride (AriceptW)

Galantamine hydrobromide (ReminylW)

Rivastigmine tartrate (ExelonW)

Active principle

Reversible and highly centrally selective inhibitor of acetylcholinesterase (AChE) with minimal effects on butrylcholinesterase (BuChE)

‘Pseudoirreversible’b centrally selective inhibitor (focus on cortex and hippocampus) of AChE and BuChE

Molecule type t1/2 Bioavailability

Piperidine-based molecule Approximately 70 h 100%; 96% bound to plasma proteins

Dual effect on the cholinergic system: selective competitive and reversible inhibitor of AchE and allosterical modulator of presynaptic nicotinic acetylcholine receptors (nAChR)a Tertiary alkaloid 7–8 h 88%; 18% bound to plasma proteins

Tmax Steady state Administration Metabolism

4h After 2–3 weeks Once daily One active metabolite (6-Odesmethyldonepezil), hepatic metabolism by CYP 450 isoenzymes 3A4 and 2D6 and glucuronidation Excreted in urine and hepatic metabolism

Elimination

1h After 2–3 days Twice daily Hepatic metabolism by CYP 450 isoenzymes 2D6 and 3A4

>90% renal elimination of metabolites

Carbamate type 0.6–2 h 36%; 40% bound to plasma proteins 1h Twice daily 95% metabolism by AChE (decarbamylation); no involvement of the CYP 450 system Renal elimination of metabolites (>90% within 24 h)

a

In vitro, galantamine binds to a site on nAChR that is different from the binding site of the natural agonist, acetylcholine (described as allosteric, meaning ‘other site’). When galantamine and acetylcholine bind simultaneously to nAChR, the response of these receptors to acetylcholine is amplified. b Pseudoirreversible means that AChE is carbamylated and after some hours hydrolyzed, that is, AChE regenerates without new synthesis.

Mostly, cognitive performance was assessed by means of the Alzheimer’s Disease Assessment Scale – Cognitive subscale (ADAS-Cog), which ranges from 0 to 70 points, 0 indicating no impairment and 70 points indicating final dementia. On average, the mean difference between the groups receiving AChE inhibitors and placebo was 3 points, which was statistically significant in favor of AChE inhibitors. This corresponds to an effect size of circa 0.5 standard deviations. In addition to the ADAS-Cog scale, the CIBIC-plus scale (Clinician’s Interview Based Impression of Change with caregiver input) was often used. This semistructured interview, performed first with the caregiver and then with the patient, provides a global clinical assessment of change. The change between baseline and endpoint assessment is marked on a 7-point Likert-like scale, where 4 represents no change compared with baseline, 1–3, different degrees of improvement, and 5–7, degrees of deterioration. On this scale, the mean difference between the active treatment and the placebo group amounted to some 0.5 points in favor of donepezil. Concerning the activities of daily living, various instruments were used in the studies. For example, in the rivastigmine and galantamine studies, improvements in the range of 3 points on the DAD (Disability Assessment for Dementia Scale, total range 0–100) and the PDS (Progressive Deterioration Scale, total range 0–100) were seen.

Identification of the Methodological Quality of the Studies

For a comprehensive assessment of methodological quality, several questions have to be answered. As most important aspects the following are considered: . Randomization/allocation concealment. Was the randomization of the patient to the treatment or placebo groups really by chance? And did neither patients nor investigators know who was randomized to which group? . Blinding. Were the investigators assessing the patients’ performance reliably blinded to the treatment of the patients? Did they not know about the side effects the patients experienced? . Handling of dropouts. Which strategies were used to include results of those patients into the analyses who dropped out of the study? . Violation of the intention-to-treat-principle. Was the intention-to-treat-principle (i.e., inclusion of all randomized patients into endpoint analyses) violated, and if yes, to what extent? The methodological quality of the studies altogether can be judged as moderate or even partly insufficient. In the first place, this is due to the fact that in most of the studies the intention-to-treat-principle was violated. Also, in more than half of the studies the dropout rates in at least one of the treatment arms amounted to 20% or more. Finally, precise

Acetylcholinesterase Inhibitors and Alzheimer’s Disease 11

information on randomization, allocation concealment, and blinding was mostly missing in the study reports. Assessment of the Relevance of the Methodological Quality for the Results

After having identified the methodological shortcomings of the studies, the crucial question is whether these shortcomings might have biased the results substantially. To date, this question is controversial. Two state institutions of health technology assessment have lately extensively evaluated the trials on AChE inhibitors. Both the British National Institute for Health and Clinical Excellence (NICE) and the German Institut fu¨r Qualita¨t und Wirtschaftlichkeit im Gesundheitswesen (Institute for Quality and Economy in Health Care) (IQWiG) confirmed the methodological shortcomings of the trials but did not assume that these shortcomings biased the results substantially. Because a definite estimation of bias caused by methodological shortcomings can only be given when original data are available, the assumption of NICE and IQWiG – who did not have the data – is not conclusive. Another question concerns the method used to account for dropouts. In most of the studies the LOCF (‘last observation carried forward’) method was used: when a patient drops out of the study prematurely, the last test results available are used as if they were results of endpoint assessment. In the case of a progressively deteriorating disease like AD, the use of LOCF can lead to better results in groups with higher dropout rates and earlier dropout time points. As in the majority of the AChE inhibitor trials, more patients dropped out of the verum groups; this may have led to better results in comparison to placebo. However, this theoretical criticism can again only be substantiated on the basis of original data. Unfortunately, the pharmaceutical companies do not disclose these data. Discussion of the External Validity of the Trials

In the majority of the RCTs, patients with othersevere ‘not controlled’ medical or psychiatric diseases, insulin-dependent diabetes mellitus, other endocrinological dysfunction, asthma, or obstructive pulmonary disease were excluded. A Canadian research group investigated to what extent a cohort of 6424 older adults who were newly dispensed donepezil in Ontario was represented in the RCTs on cholinesterase inhibitors. They found that 51% to 78% of the patients would not have been included in the trials. This means that, strictly speaking,

the validity of the results of these trials is limited to some half of the patients. A second point concerning the external validity of the trials is their short duration, as most of the RCTs do not exceed 6 months. In consequence, there is hardly any RCT-based evidence for treatment lasting for more than 6 months. Determination of the Clinical Relevance of Effects and Side Effects as Measured in the Trials

What does a mean group difference of 3 points on a 70 points cognitive scale mean? The answer is unclear. A common answer given is that a US Food and Drug Administration (FDA) panel defined a change of 4 points on the ADAS-Cog as clinically relevant. But irrespective of the scientific basis of this consensus definition, these 4 points apply to changes in individuals and not in groups. The answer on the question of clinical relevance could be clearer if the distribution of changes had been published. Assuming a normal distribution, IQWiG calculated that the difference in success rates between the cholinesterase inhibitor and placebo groups amounts to 13–16% if success is defined as an individual improvement of 4 points. This means that at least six to eight patients have to be treated with the active drug in order to see success in cognition in one patient. Concerning activities of daily living, the number needed to treat amounts to 13–14. But these calculations have to be interpreted with caution, as they are based on the assumption of a normal distribution of the results. The leading question concerning clinical relevance is the impact of treatment on life quality of patients and caregivers. Does the one out of six to eight patients who scores 4 points better on the ADASCog experience any benefit in everyday life? There is a consensus that the data from the RCTs are not sufficient (either results are inconsistent or life quality was not evaluated) to ascertain a benefit concerning life quality of patients and caregivers. The same applies to institutionalization and a desirable disease-modifying effect: There is no evidence from RCTs that cholinesterase inhibitors delay institutionalization of patients or that they modify disease progression. The question of side effects is also not controversial. AChE inhibitors have side effects: nausea, vomiting, diarrhea, weight loss, and agitation are typical. They occur in 5–10% of patients taking donepezil, 5–20% of patients taking galantamine, and 10–40% of patients taking rivastigmine. They can occur just in the titration phase of the drug, but can also persist and are the main reason for dropout from the

12 Acetylcholinesterase Inhibitors and Alzheimer’s Disease

trials and for discontinuation of therapy in clinical practice. Comprehension of the Clinical Experience of Patients, Caregivers, and Health Personnel

Clinicians state that they see a clear benefit in 5% to 15% of their patients, but it has not been possible to identify these responders before treatment. Here again, hypotheses could be deduced from the studies that already exist, but the pharmaceutical companies do not make the original data available. Another point is that some clinicians or caregivers point out that they see improvement in areas that are difficult to measure in standardized tests. For example, in their opinion the patient appears to be better regulated. Such an observation is in fact difficult to measure and to quantify. One of the latest studies on galantamine tried to overcome this dilemma and measured treatment outcomes by means of the Goal Attainment Scaling (GAS). The GAS is a personalized outcome measure in which people set goals according to their own needs and define improved or worsened states in their own words. After 16 weeks, at the end of double blind treatment, clinicians reported statistically significant changes on the GAS, whereas patients/caregivers did not. Even though the data do not support the assumption that the GAS is more sensitive in detecting treatment effects than other commonly used instruments in cholinesterase inhibitor trials, this instrument highlights the views of patients and caregivers as to whether treatment is seen as meaningful. Further research on this subject is needed.

Consequences of Unclear Clinical Benefit Reflecting the controversial estimation of clinical benefit, national regulations concerning prescription and treatment documentation differ greatly. For example, in Germany anybody with mild to moderate AD can be treated with cholinesterase inhibitors as long as the treating doctor prescribes the drug, without any further restrictions. In many countries (e.g., Belgium, France, the UK, Australia), in contrast, there are clear rules concerning initiation, evaluation, and termination of therapy. In the Netherlands, donepezil and galantamine (and memantine) are not covered by the statutory health insurance – only rivastigmine is – on the basis of a European Unionwide admission procedure. In Belgium, a patient registry is to be established to follow these patients. Data from such registries could be used to analyze prescription patterns and effects in the great variety of patients treated under real life conditions.

Imprecise Cost Estimates Costs for the drugs in Germany amount to $1600 to $2000 per year. Not included are all other costs incurred, for example, costs of doctor consultations to get the prescription or costs of diagnosis and treatment of side effects. The data basis for cost estimates in relation to efficacy is extremely arguable. In their final appraisal document, the NICE reports amounts of $39 000 to $260 000 per quality adjusted life year gained. A huge range is also reported for the costs of gaining an additional year in a nonsevere state: $2200 up to $26 000. All these cost calculations are difficult to follow.

Summary . The appropriateness of including RCTs only in assessments of efficacy of cholinesterase inhibitors is questioned, but there is a broad consensus that RCTs constitute the basis. . The published RCTs on cholinesterase inhibitors have methodological shortcomings. . The relevance of the methodological shortcomings is valued differently. . Without original data it is not possible to prove whether the methodological shortcomings substantially biased the results or not. . If the methodological shortcomings are ignored, the study results show small differences between the cholinesterase inhibitor and placebo groups. The clinical relevance of these differences is controversial. . There is no evidence from RCTs concerning benefits on life quality of patients and caregivers and delay of institutionalization. . Due to broad exclusion criteria and the duration of the trials mostly not exceeding 6 months, the external validity of the trials is rather low. . Cholinesterase inhibitors have side effects. . No disease-modifying effect has been able to be measured. . Trial data from the industry should be accessible for research purposes. . Further research directed at improved pharmacological therapy for patients with AD is needed.

See also: Acetylcholine Neurotransmission in CNS; Acetylcholinesterase; Aging of the Brain and Alzheimer’s Disease; Alzheimer’s Disease: An Overview; Basal Forebrain and Memory; Cholinergic System; Cholinergic Pathways in CNS; Cholinergic Neurotransmission in the Autonomic and Somatic Motor Nervous System; Cholinergic System Imaging in the Healthy Aging Process and Alzheimer Disease; Nicotinic Acetylcholine

Acetylcholinesterase Inhibitors and Alzheimer’s Disease 13 Receptors; Sleep–Wake State Regulation by Acetylcholine.

Further Reading Burns A and O’Brien J on behalf of the BAP Dementia Consensus Group (2006) Clinical practice with anti-dementia drugs: A consensus statement from British Association for Psychopharmacology. Journal of Psychopharmacology 20(6): 732–755. Gill SS, Bronskill SE, Mamdani M, et al. (2004) Representation of patients with dementia in clinical trials of donepezil. Canadian Journal of Clinical Pharmacology 11(2): e274–e285. IQWiG (2006) Cholinesterasehemmer bei Alzheimer Demenz [Cholinesterase inhibitors in Alzheimer’s disease]. Vorbericht

A05/19-A. Ko¨ln: Institut fu¨r Qualita¨t und Wirtschaftlichkeit im Gesundheitswesen (IQWiG). Kaduszkiewicz H, Zimmermann T, Beck-Bornholdt H-P, and van den Bussche H (2005) Cholinesterase inhibitors for patients with Alzheimer’s disease: Systematic review of randomized clinical trials. British Medical Journal 331: 321–327. National Institute for Health and Clinical Excellence (NICE) (2006) Alzheimer’s disease – Donepezil, rivastigmine, galantamine and memantine (review) – Final appraisal document. 26 May 2006. Rockwood K, Fay S, Song X, MacKnight C, Gorman M, and Video-Imaging Synthesis of Treating Alzheimer’s Disease (VISTA) Investigators (2006) Attainment of treatment goals by people with Alzheimer’s disease receiving galantamine: A randomized controlled trial. Canadian Medical Association Journal 174(8): 1099–1105.