What_happens_when_donepezil_is_suddenly.pdf

INTERNATIONAL JOURNAL OF GERIATRIC PSYCHIATRY Int J Geriatr Psychiatry 2003; 18: 988–993. Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/gps.995

What happens when donepezil is suddenly withdrawn? An open label trial in dementia with Lewy bodies and Parkinson’s disease with dementia Thaı´s S. C. Minett, Alan Thomas, Lucy M. Wilkinson, Sarah L. Daniel, Jonathan Sanders, Jonathan Richardson, Elizabeth Littlewood, Pat Myint, Jane Newby and Ian G. McKeith* Institute for Ageing and Health, Wolfson Research Centre, Newcastle General Hospital, Newcastle upon Tyne, UK

SUMMARY Background This open label study was designed to assess the effects of donepezil treatment, its withdrawal and subsequent recommencement on cognitive functioning, behaviour and parkinsonian symptoms in patients with probable dementia with Lewy bodies (DLB) and with Parkinson’s disease who subsequently developed dementia (PDD). Methods Eight patients with DLB and 11 with PDD were treated with up to 10mg of donepezil daily for 20 weeks followed by a 6-week withdrawal period. The primary outcome measures were the Mini-Mental State Examination (MMSE), the total Neuropsychiatric Inventory (NPI) and the Unified Parkinson’s Disease Rating Scale III. Testing was conducted before dosing, at week 20, at a withdrawal visit and 3 months after recommencement on donepezil. Results Patients with DLB and PDD showed a significant improvement in cognition with treatment, loss of this improvement on withdrawal and restoration of treatment gains on recommencement. Both groups also demonstrated favourable, behavioural changes with treatment, PDD patients in particular deteriorating significantly after withdrawal. The only NPI symptom domain that showed a consistent significant response to both treatment (positive) and withdrawal (negative) was hallucinations. The medication was well tolerated and parkinsonian features did not alter significantly over the testing sessions. Conclusions Our results suggest that treatment with donepezil improves cognition and hallucinations without increasing parkinsonian symptoms, and its sudden withdrawal is usually detrimental, producing acute cognitive and behavioural decline. Although recommencement on donepezil appears to reverse this deterioration we do not advise its abrupt discontinuation in this population. Copyright # 2003 John Wiley & Sons, Ltd. key words — dementia; Lewy bodies; Parkinson’s disease; behaviour; cognition; psychopathology; withdrawal; cholinesterase inhibitors; donepezil; clinical trial

INTRODUCTION Parkinson’s disease (PD) and dementia with Lewy bodies (DLB) (McKeith et al., 1996) are clinical terms for two overlapping presentations of Lewy body disease. The definitive diagnosis of both requires the pathological identification of Lewy bodies. Clinical *Correspondence to: Prof. I. G. McKeith, Institute for Ageing and Health, Wolfson Research Centre, Newcastle General Hospital, Westgate Road, Newcastle upon Tyne, NE4 6BE, UK. Tel: þ44 191 256 3018. Fax: þ44 191 219 5071. E-mail: [email protected] Contract/grant sponsor: FAPESP; contract/grant number: 99/ 04928-9. Contract/grant sponsor: CAPES; contract/grant number: BEX 0096/01-6. Copyright # 2003 John Wiley & Sons, Ltd.

diagnosis of probable DLB is made in the presence of a dementia syndrome characterised by progressive cognitive decline, and two of the following core features: fluctuating cognition with pronounced variations in attention and alertness; recurrent visual hallucinations, and spontaneous motor features of parkinsonism. These criteria are potentially applicable to patients with idiopathic PD who subsequently develop dementia (PDD). If dementia occurs within a year of the onset of parkinsonism, patients should be diagnosed as having DLB. However, if the dementia starts more than one year after parkinsonism, the diagnosis should be PD þ DLB (McKeith et al., 1996). The particular importance of these syndromes, compared with other dementia types, relates to how Received 23 April 2003 Accepted 15 July 2003

sudden withdrawn of donepezil they are managed. Treating behavioural disturbances in DLB with neuroleptics often causes a marked worsening of extrapyramidal symptoms, and this can be life-threatening. Neuroleptic sensitivity reactions are associated with a poor outcome and two- to three-fold increased mortality (Barber et al., 2001). Decrements in choline acetyltransferase (ChAT) activity have been found in the basal forebrain in AD and Lewy body disease, especially in the nucleus basalis of Meynert (Tiraboschi et al., 2000). There is increasing evidence that DLB is associated with a greater neocortical, cholinergic deficit than that observed in AD; particularly in the frontal, temporal and parietal cortices (Perry et al., 1993). This may contribute to the higher prevalence of psychotic symptoms in DLB. Increased post-synaptic cortical muscarinic receptor binding occurs in DLB (Perry et al., 1993), suggesting that cholinergic replacement therapy should be especially effective. Preliminary clinical data using cholinesterase inhibitors (ChE-I) in DLB, has shown improvements in cognitive and behavioural symptoms. Among the three (ChE-I) currently in use (rivastigmine, donepezil and galantamine), only rivastigmine has been used in a randomised, double-blind, placebo-controlled trial in patients with DLB: a 20 week study which included 120 patients from the UK, Spain and Italy (McKeith et al., 2000a; Wesnes et al., 2002). This study showed that rivastigmine substantially improved (by at least 30%) behavioural symptoms in 63.4% of patients, whereas only 30.0% of patients taking placebo had the same positive response. The cognitive symptoms also showed a pattern in favour of rivastigmine (Wesnes et al., 2002). Data on donepezil-treated patients with DLB or PDD is also available as case studies (Shea et al., 1998; Lanctot and Herrmann, 2000) and small trials (Samuel et al., 2000; Aarsland et al., 2002; Bergman and Lerner, 2002). All of these support the same trend. This paper reports the findings of an open label trial, which had the following objectives: 1. To compare the profile of response of DLB and PDD patients to treatment with donepezil; 2. To assess the effects of donepezil withdrawal; 3. To determine whether recommencement of donepezil reversed any changes after withdrawal and; 4. To examine the efficacy and tolerability of donepezil in patients with Lewy body disease. METHOD This was an exploratory, single centre, open label study, recruiting patients over 12 months. Patients Copyright # 2003 John Wiley & Sons, Ltd.

989

referred from the Newcastle and Gateshead Old Age Psychiatry and Neurology Outpatient Departments were eligible if they either had a primary diagnosis of probable DLB or PDD. The consensus clinical diagnostic criteria (McKeith et al., 1996) and the Parkinson’s disease brain-bank criteria (Gibb and Lees, 1988) were used to diagnose DLB and PD respectively. All the PDD patients had a dementia syndrome with features meeting the DLB profile. Patients were excluded if they had severe gastrointestinal, renal or liver disease, a history of cardiac bradyarrhythmias, asthma or bladder outflow obstruction, or recent history of cerebrovascular disease. Patients were also excluded if they were taking cholinergic, anti-cholinergic, non-steroidal anti-inflammatory medication or neuroleptics. Informants with regular day-to-day contact with the patients were required to attend each visit to complete assessment schedules and oversee the administration of medication. Ethical approval was granted for this study. Eligible patients received information about the study, and those who gave informed consent were entered. At baseline, the full assessment battery (see below) was completed. Donepezil 5 mg was then prescribed to be taken orally at night. Patients were titrated to a dose of 10 mg nocte four weeks after unless they were already experiencing significant side-effects at the lower dose. The dose could be reduced back to 5 mg at any time during the treatment period. Patients were contacted by telephone one week after baseline and were reassessed at week 4 to assess tolerability and to record any adverse events, particularly nausea, vomiting, diarrhoea, headache, cramps, confusion, and agitation. At week 20, patients completed their efficacy assessment and medication was stopped immediately without any tapered reduction. The withdrawal period was for 6 weeks following which patients were again assessed; in some cases where there was a rapid deterioration leading to distress to patients and carers the withdrawal phase was shorter. Patients were finally assessed after 3 months recommencement on donepezil. Cognitive impairment was assessed using the MiniMental State Examination (MMSE) (Folstein et al., 1975), behavioural symptoms were measured using the Neuropsychiatric Inventory (NPI) 12 items (Cummings, 1997), and parkinsonian symptoms were assessed using the Unified Parkinson’s Disease Rating Scale III (UPDRS III) (Fahn and Elton, 1987). Mean values between patient groups were compared using independent sample t-tests (t). Mean scores pre, during, and post treatment were compared using paired sample t-tests (t). In view of the Int J Geriatr Psychiatry 2003; 18: 988–993.

990

t. s. c. minett ET AL.

non-normal distribution of the NPI data, the Wilcoxon rank sum test (Z ) or the Wilcoxon signed ranks test (Z) was used throughout. A p-value of <0.05 was considered to indicate statistical significance; all tests were two-tailed. Patients for whom relevant data was missing were excluded from the analyses. All statistical analysis was performed with the statistical package SPSS for Windows (version 10). RESULTS Twenty-four patients were recruited, nine with a primary diagnosis of DLB, and 15 with PDD. Among all patients at baseline, 18 had all three core diagnostic symptoms and six had two. Twenty-three had visual hallucinations, 20 had fluctuations and 22 had spontaneous parkinsonism. Sixteen patients had one or two supportive features of DLB and three had three or four. The most frequent supportive features were systemised delusions and repeated falls, presented in 12 and 11 patients, respectively. There were no statistically significant differences between patients with DLB and with PDD at baseline in terms of mean MMSE scores (15.3 vs 18.2, t (22) ¼
1.15, p ¼ 0.261, 95% Confidence Intervals (CI)
7.9–3.0), mean UPDRS III scores (26.9 vs 34.2, t (22) ¼
1.70, p ¼ 0.104, 95% CI
15.7–3.0) and the total NPI scores (Z ¼
0.4, p ¼ 0.679). Adverse effects The total duration of treatment was 20 weeks. Of the 24 patients who started on treatment, five discontinued in the course of the trial: four left due to side effects including nausea, vomiting, diarrhoea, confusion, increased hallucinations, insomnia, cramps, falls and worsening of parkinsonism. One patient was admitted to hospital with a gastric ulcer and empyema of the gall bladder. The mean time on treatment for those who discontinued was five weeks (range ¼ 0.7 to 19). Otherwise, the adverse effects were not severe, and the most common are presented in Table 1. Nineteen patients (eight DLB and 11 PDD) completed the study and were included in further analyses. Response to treatment MMSE. The mean MMSE scores at week 20 were significantly higher than at baseline, indicating improvement under treatment for both groups of patients: DLB (19.9 vs 15.8, mean difference ¼ 4.1, t (7) ¼ 3.7, p ¼ 0.007, 95% CI 1.5–6.7) and PDD Copyright # 2003 John Wiley & Sons, Ltd.

Table 1. Adverse effects which occurred in at least 15% of patients taking donepezil Adverse effect

Number of patients (percent)

Dizziness Insomnia Polyuria Constipation/diarrhoea Cramps Hypersalivation Nausea/vomiting Tremor

4 (16.7) 5 (20.8) 5 (20.8) 6 (25.0) 6 (25.0) 7 (29.2) 7 (29.2) 7 (29.2)

Figure 1. Mean MMSE scores in DLB and PD þ DLB patients from baseline to 3 month extension. Error bars depict standard deviations. * ¼ significant at 0.05 level compared to baseline, ** ¼ significant at 0.05 level compared to week 20. þ ¼ DLB group (n ¼ 5), þþ ¼ PD þ DLB group (n ¼ 10)

(21.3 vs 17.5, mean difference ¼ 3.8, t (10) ¼ 4.2, p ¼ 0.002, 95% CI 1.7–5.7) (Figure 1). NPI. At baseline, only one patient scored zero in the total NPI. Even though comparisons of the total NPI scores at week 20 with baseline did not show significant reductions for either group of patients, six out of eight patients with DLB and four of 11 with PDD, experienced clinically relevant behavioural improvements (defined by >30% reduction in their total NPI scores). The NPI symptom domains improving significantly with treatment in patients with DLB included hallucinations (Z ¼
2.0, p ¼ 0.041), depression (Z ¼
2.0, p ¼ 0.046) and anxiety (Z ¼
2.2, p ¼ 0.026), and within the PD þ DLB group, were delusions (Z ¼
2.0, p ¼ 0.042), aberrant motor behaviour (Z ¼
2.0, p ¼ 0.042) and hallucinations (Z ¼
2.1, p ¼ 0.029). The medians of the total NPI scores are presented in Figure 2. Int J Geriatr Psychiatry 2003; 18: 988–993.

991

sudden withdrawn of donepezil

and hallucinations: Z ¼
1.1, p ¼ 0.273; DLB: hallucinations: Z ¼
0.4, p ¼ 0.671). UPDRS III. The mean UPDRS III scores for both groups of patients did not alter significantly from baseline. Response to recommencement of donepezil

Figure 2. Median of total NPI scores in DLB and PD þ DLB patients from baseline to 3 month extension. * ¼ significant at 0.05 level compared to week 20. þ ¼ DLB group (n ¼ 5), þþ ¼ PD þ DLB group (n ¼ 10)

Response to withdrawal Of the 19 patients, ten (four DLB and six PDD) asked to restart treatment before the 6 week withdrawal was completed because of marked clinical deterioration. There was no statistically significant difference in the mean length of the withdrawal period between patients with DLB and with PDD (34.3 vs 37.6 days, t (17) ¼
0.43, p ¼ 0.673, 95% CI
20.0–13.2). MMSE. Both groups of patients showed a marked deterioration in the MMSE scores after withdrawal when compared to treatment (Figure 1). However, when compared to their baseline scores there were no significant changes (difference in baseline and withdrawal mean MMSE scores for DLB patients ¼ 1.1, t (7) ¼
1.32, p ¼ 0.229, 95% CI ¼
3.1–0.9; or for PDD patients ¼ 1.1, t (10) ¼ 1.3, p ¼ 0.221, 95% CI ¼
0.8–2.9) (Figure 1). NPI. Only patients with PDD showed a statistically significant worsening in the total NPI after withdrawal (Z ¼
2.6, p ¼ 0.008). However, this worsening failed to reach significance when compared to baseline (Z ¼
1.6, p ¼ 0.110); see Figure 2. The symptom domains which deteriorated significantly compared with week 20 in patients with PDD, were delusions (Z ¼
2.5, p ¼ 0.013), irritability (Z ¼
2.0, p ¼ 0.043) and hallucinations (Z ¼
2.1, p ¼ 0.034); and in patients with DLB, hallucinations (Z ¼
2.0, p ¼ 0.046). However, these symptoms did not worsen significantly when comparing baseline with withdrawal visit (PDD: delusions: Z ¼
1.7, p ¼ 0.089; irritability: Z ¼
1.2, p ¼ 0.236; Copyright # 2003 John Wiley & Sons, Ltd.

MMSE. Scores were available for five of eight DLB and 10 of 11 PDD patients at 3 months after recommencement of treatment. One patient died, two withdrew and one did not have an MMSE. There were no significant differences between week 20 and the 3 month recommencement (Z ¼ 0.32, p ¼ 0.75) for the combined group or for DLB (Z ¼ 0.95, p ¼ 0.343) and PDD (Z ¼ 0.43, p ¼ 0.67) groups separately. NPI. The same 15 subjects completed 3 month NPI ratings and there were again no differences between week 20 and the end of this recommencement on treatment period: combined group, Z ¼ 0.32, p ¼ 0.753; DLB group, Z ¼ 0.14, p ¼ 0.893; PDD group, Z ¼ 0.63, p ¼ 0.528. UPDRS III. The mean UPDRS III scores for both groups of patients did not alter significantly between week 20 and the 3 month recommencement period. DISCUSSION The main findings of this open label trial were confirmation of the efficacy of cholinesterase inhibitor treatment in DLB, demonstration of similar efficacy in PDD, no evidence of worsening of motor parkinsonism, and evidence that withdrawal of treatment leads to rapid clinical deterioration but that recommencement appears to restore the gains of treatment. In the absence of a placebo control, these data must be regarded cautiously but we believe they are suggestive of an important aspect of clinical management of LB disease patients that requires further clarification. The efficacy of donepezil in DLB is consistent with the only randomised controlled trial of cholinesterase inhibitor treatment in patients with DLB, which used rivastigmine (McKeith et al., 2000a). It is well established that cholinesterase inhibitors improve cognitive symptoms in AD and others have reported significant benefits in patients with DLB and PDD (McKeith et al., 2000b; Bergman and Lerner, 2002). Our study showed that donepezil significantly improved cognition in patients with DLB and PDD, Int J Geriatr Psychiatry 2003; 18: 988–993.

992

t. s. c. minett ET AL.

with a mean improvement of 4 points on the MMSE. In the rivastigmine trial apathy was the neuropsychiatric disturbance which most benefited from treatment, whereas in ours it was hallucinations in both groups of patients. Hallucinations and sleep disturbance were the psychiatric features which improved most in PDD patients treated with rivastigmine in an open label trial (Reading et al., 2001). Donepezil was generally well tolerated in our study. The frequency and severity of adverse effects was not worse than in patients with AD (Rogers and Friedhoff, 1996; Greenberg et al., 2000). We found no indication that this treatment worsens parkinsonian symptoms. One patient discontinued with increased bradykinesia, but the main reasons for his dropout were agitation and confusion. A crossover placebo controlled trial using donepezil in patients with PDD corroborates our findings regarding parkinsonian symptoms (Aarsland et al., 2002). In that study, UPDRS motor scale scores showed no deterioration after 10 weeks of treatment when comparing with placebo or with their pre-treatment levels. Shea et al. (1998) reported worsening of parkinsonism in two out of nine patients with DBL treated with donepezil, however, they did not perform any quantitative measurement of parkinsonian features, invalidating statistical analysis. All other studies assessing parkinsonism in DLB or PDD have found either no change or improvement with ChE-I treatment (Lanctot and Herrmann, 2000; McKeith et al., 2000b; Samuel et al., 2000; Reading et al., 2001). The abrupt withdrawal of donepezil, i.e. without downward dose titration, was detrimental as it produced a substantial decline in cognition and behavioural symptoms for both groups of patients. Similar results were shown in the crossover study in patients with PDD (Aarsland et al., 2002): in the group receiving donepezil first, the MMSE score six weeks after crossover to placebo was similar to baseline. Deteriorating cognition and behaviour in patients with PDD or DLB has also been reported following abrupt rivastigmine withdrawal (McKeith et al., 2000a; Reading et al., 2001; Wesnes et al., 2002). Just over half of our patients deteriorated rapidly and were unwilling to wait six weeks before reinstatement of treatment. Since they lost their clinical improvements with a return to baseline values we suggest that withdrawal of ChE-I in this group is likely to be detrimental and that patients showing good symptomatic response should be continued on long term treatment (Grace et al., 2001). If treatment is stopped in a patient who has previously responded but is now deteriorating, or because of adverse effects, Copyright # 2003 John Wiley & Sons, Ltd.

KEY POINTS *

*

*

*

Cholinesterase inhibitors including donepezil improve cognitive and neuropsychiatric symptoms in DLB and PDD. In this study sudden withdrawal of donepezil was usually detrimental, producing acute cognitive and behavioural decline. Although recommencement on donepezil reversed this deterioration we do not advise its abrupt discontinuation in DLB and PDD patients. Our study was limited by absence of a placebo control group and therefore needs further confirmation.

gradual dose reduction may ameliorate the rate of associated decline. Since patients in our study who recommenced treatment after withdrawal regained the benefits they had previously derived within 3 months, a cautious withdrawal seems a reasonable course in cases where efficacy is unclear since there is no evidence for long lasting clinical deterioration. This is an open label trial and therefore there are significant limitations to the conclusions that can be drawn from it alone. The positive benefits of ChE-I in DLB and PDD are, however, leading to their increased clinical use in these diagnoses and we recommend caution in sudden withdrawal of treatment. Randomised withdrawal studies would clarify the situation but may be ethically problematic and difficult to conduct in practice. ACKNOWLEDGEMENT TSC Minett was supported by grants from FAPESP (99/04928-9) and CAPES (BEX 0096/01-6). REFERENCES Aarsland D, Laake K, Larsen JP, Janvin C. 2002. Donepezil for cognitive impairment in Parkinson’s disease: a randomised controlled study. J Neurol Neurosurg Psychiatry 72: 708–712. Barber R, Panikkar A, McKeith IG. 2001. Dementia with Lewy bodies: diagnosis and management. Int J Geriatr Psychiatry 16: S12–S18. Bergman J, Lerner V. 2002. Successful use of donepezil for the treatment of psychotic symptoms in patients with Parkinson’s disease. Clin Neuropharmacol 25: 107–110. Cummings JL. 1997. The Neuropsychiatric Inventory: assessing psychopathology in dementia patients. Neurology 48: S10–S16. Fahn S, Elton R. 1987. Unified Parkinson’s Disease Rating Scale. In Recent Developments in Parkinson’s Disease, Fahn S, Marsden CD, Calne D (eds). Macmillan: New York; 153–163. Int J Geriatr Psychiatry 2003; 18: 988–993.

sudden withdrawn of donepezil Folstein MF, Folstein FE, McHugh PR. 1975. ‘Mini-Mental State’: a practical method for grading the cognitive state of patients for the clinician. J Psychiat Res 12: 189–201. Gibb WR, Lees AJ. 1988. The relevance of the Lewy body to the pathogenesis of idiopathic Parkinson’s disease. J Neurol Neurosurg Psychiatry 51: 745–752. Grace J, Daniel S, Stevens T, et al. 2001. Long-Term use of rivastigmine in patients with dementia with Lewy bodies: an openlabel trial. Int Psychogeriatr 13: 199–205. Greenberg S, Tennis MK, Brown LB, et al. 2000. Donepezil Therapy in clinical practice: a randomized crossover study. Arch Neurol 57: 94–99. Lanctot KL, Herrmann N. 2000. Donepezil for behavioural disorders associated with Lewy bodies: a case series. Int J Geriatr Psychiatry 15: 338–345. McKeith I, Del Ser T, Spano P, et al. 2000a. Efficacy of rivastigmine in dementia with Lewy bodies: a randomised, double-blind, placebo-controlled international study. Lancet 356: 2031–2036. McKeith IG, Galasko D, Kosaka K, et al. 1996. Consensus guidelines for the clinical and pathological diagnosis of dementia with Lewy bodies (DLB): report of the consortium on DLB international workshop. Neurology 47: 1113–1124. McKeith IG, Grace JB, Walker Z, et al. 2000b. Rivastigmine in the treatment of dementia with Lewy Bodies: preliminary findings from an open trial. Int J Geriat Psychiatry 15: 387–392.

Copyright # 2003 John Wiley & Sons, Ltd.

993

Perry EK, Irving D, Kerwin JM, et al. 1993. Cholinergic transmitter and neurotrophic activities in Lewy body dementia: similarity to Parkinson’s and distinction from Alzheimer disease. Alzheimer Dis Assoc Disord 7: 69–79. Reading PJ, Luce AK, McKeith IG. 2001. Rivastigmine in the treatment of parkinsonian psychosis and cognitive impairment: preliminary findings from an open trial. Mov Disord 16: 1171–1174. Rogers SL, Friedhoff LT. 1996. The efficacy and safety of donepezil in patients with Alzheimer’s disease: results of a US Multicentre, randomized, double-blind, placebo-controlled trial. The donepezil study group. Dementia 7: 293–303. Samuel W, Caligiuri M, Galasko D, et al. 2000. Better cognitive and psychopathologic response to donepezil in patients prospectively diagnosed as dementia with Lewy bodies: a preliminary study. Int J Geriatr Psychiatry 15: 794–802. Shea C, MacKnight C, Rockwood K. 1998. Donepezil for treatment of dementia with Lewy bodies: a case series of nine patients. Int Psychogeriatr 10: 229–238. Tiraboschi P, Hansen LA, Alford M, et al. 2000. Cholinergic dysfunction in diseases with Lewy bodies. Neurology 54: 407–411. Wesnes KA, McKeith IG, Ferrara R, et al. 2002. Effects of rivastigmine on cognitive function in dementia with Lewy bodies: a randomised placebo-controlled international study using the cognitive drug research computerised assessment system. Dement Geriatr Cogn Disord 13: 183–192.

Int J Geriatr Psychiatry 2003; 18: 988–993.