The long-term efficacy and tolerability of donepezil in patients with vascular dementia

RESEARCH ARTICLE The long-term efficacy and tolerability of donepezil in patients with vascular dementia David Wilkinson 1, Gustavo Róman 2, Stephen Salloway 3, Jane Hecker 4, Karyn Boundy 5, Dinesh Kumar 6, Holly Posner 7 and Rachel Schindler 7 1 Memory Assessment and Research Centre, Moorgreen Hospital, Southampton, UK 2 University of Texas Medical School at San Antonio, San Antonio, TX, USA 3 Memory and Aging Program, Butler Hospital, Providence, RI, USA 4 Memory Disorders Study Unit, Repatriation General Hospital, Daw Park, Australia 5 Department of Neurology, Queen Elizabeth Hospital, Woodville, Australia 6 Eisai Inc., Woodcliff Lake, NJ, USA 7 Pfizer Inc., New York, NY, USA Correspondence to: D. Wilkinson, E-mail: David.Wilkinson@hantspt-sw.nhs.uk Objective: To determine the long-term tolerability and efficacy of donepezil in patients with vascular dementia (VaD). Methods: International, multicentre, open-label, 30-week extension study of two 24-week, randomised, double-blind, placebo-controlled studies. Participants were ambulatory adults (59% female; mean age, 74.7 0.3) with a diagnosis of possible or probable VaD and without a diagnosis of Alzheimer s disease, who were medically stable and had completed one of two double-blind studies. All patients received donepezil 5 mg/day for the first 6 weeks, then 10 mg/day (clinician approval required). Assessments were performed at week 6 and every 12 weeks thereafter. The main outcome measure was the Alzheimer s disease Assessment Scale-cognitive subscale (ADAS-cog). Safety/tolerability measures included adverse events (AEs) and physical and laboratory evaluations. Results: Of 1219 eligible patients, 885 (72.6%) were enrolled, of which 707 (79.9%) completed the study; 127 (14.4%) patients discontinued due to AEs. A mean reduction (0.6 1.15 points) from double-blind study baseline score to week 54 (end of open-label study) on the ADAS-cog was observed for patients who received donepezil continuously for 54 weeks. ADAS-cog scores remained stable in the group that initiated donepezil treatment during the extension study. Most common donepezil-related AEs were nausea (occurring in 5.3%) and diarrhoea (8.8%); no unexpected AEs attributable to donepezil occurred. Conclusion: These data suggest that donepezil improves cognition for up to 54 weeks in patients with VaD. Patients initiating donepezil in this extension study did not perform as well on the primary outcome measure as those initiating donepezil in the double-blind study. Copyright # 2009 John Wiley & Sons, Ltd. Key words: vascular dementia; donepezil; cholinesterase inhibitor; long-term trial; safety; tolerability; efficacy History: Received 20 January 2009; Accepted 21 May 2009; Published online 21 July 2009 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/gps.2340 Introduction Vascular dementia (VaD), one of the most common forms of dementia (Dubois and Hebert, 2001), results from ischaemic injury to brain regions serving memory, cognition and behaviour. In many countries throughout the world, there is currently no approved therapy for the symptomatic treatment of VaD (Doody et al., 2001b), and treatment strategies generally focus on the control of underlying cardiovascular risk factors such as the treatment of hypertension, correction of arrhythmias, smoking cessation, management of

306 D. Wilkinson et al. hyperlipidaemia and control of metabolic disorders such as diabetes. Given that brain lesions in patients with VaD may produce cholinergic dysfunction similar to that seen in patients with Alzheimer s disease (AD) (Tohgi et al., 1996), donepezil, a cholinesterase inhibitor (ChEI) approved for the treatment of AD, was investigated for the treatment of VaD. There have been two published 24-week, double-blind, placebo-controlled trials of donepezil in patients with possible or probable VaD (Black et al., 2003; Wilkinson et al., 2003), diagnosed clinically and radiologically using NINDS-AIREN criteria (Roman et al., 2005). A combined analysis of these studies (N ¼ 1219) showed that donepezil provided significant improvements at study end compared with placebo for cognition at both 5 and 10 mg/day doses, and for global function at the 5 mg/ day dose; donepezil was also well tolerated (Roman et al., 2005). A third, currently unpublished, trial of donepezil for VaD has been conducted, which was very similar in scope and structure to the two published trials, except that screening computed tomography or magnetic resonance imaging scans were reviewed centrally (Roman et al., 2006). As a result of available trial data, donepezil is now licensed for the treatment of VaD in Romania, India, South Korea, Thailand, Vietnam, New Zealand and the Philippines. The objective of the current study, an open-label extension of the original double-blind studies (Black et al., 2003; Wilkinson et al., 2003), was to determine the long-term safety and sustainability of efficacy in donepezil-treated patients, and to compare the effect of a delayed start to therapy in the placebo patients with effects in patients continually treated with donepezil. Methods Study design This was an open-label, 30-week extension study of two 24-week, randomised, double-blind, placebo-controlled, parallel-group studies in patients with possible or probable VaD. The study designs and results for the original trials have been published previously (Black et al., 2003; Wilkinson et al., 2003). Study sites conducted at 100 sites in the US, UK and Europe. For a subset of sites in the United Kingdom, Ireland, Germany and Australia, the study was extended for an additional 36 weeks (total extension of 66 weeks); this was for investigators who specifically made the request for their patient(s). Due to a small number of evaluable patients entering the additional 36-week extension phase ( 77 patients per original treatment group), as well as the potential for selection bias, efficacy data beyond the original 30-week extension were not statistically viable and are not presented in the current report. Safety findings from the additional 36- week extension are reported. Study entry criteria All patients who completed the double-blind studies were eligible for entry into the open-label study. Entry criteria for the double-blind studies have been described previously (Black et al., 2003; Wilkinson et al., 2003; Roman et al., 2005). Briefly, men and women aged at least 40 years with a diagnosis of probable or possible VaD of at least 3 months duration, Mini- Mental State Examination (MMSE) scores between 10 and 26, and with clinical and radiological evidence of cerebrovascular disease were included. Patients with hypertension, diabetes, cardiac disease or stroke were included provided these conditions had been stable for at least 3 months. Patients with AD or other neurodegenerative disorders were excluded, as were those with major psychiatric disorders or a history of drug or alcohol abuse. Most concomitant medications were permitted during the study, with the exception of other ChEIs and anticholinergics. All patients (or their legal representative) and their caregivers provided voluntary, written informed consent for participation in the study. This study was conducted in accordance with the ethical principles of the Declaration of Helsinki, as amended, and in accordance with local laws and regulations, whichever afforded the greater protection for the participant. Protocol In this open-label extension, all patients received donepezil 5 mg/day for the first 6 weeks and 10 mg/day thereafter, if approved by a clinician (irrespective of their assigned group during the double-blind studies), for a total of 30 weeks. Patients were assessed at week 6 and thereafter at 12-week intervals. In the subset of patients participating in the additional study extension, donepezil was continued at the same dosage and assessments were performed at weeks 42, 54 and 66. Vital signs and adverse events (AEs) were monitored at each assessment. A physical and neurological examination was also conducted and laboratory tests (clinical chemistry, haematology and urinalysis) performed.

Long-term use of donepezil for vascular dementia 307 Outcome measures The primary efficacy measure was the Alzheimer s disease Assessment Scale- cognitive subscale (ADAScog), a measure of cognitive function. Secondary efficacy outcome measures were the Clinical Dementia Rating-Sum of the Boxes (CDR-SB), a measure of overall disease severity; the MMSE, a second measure of cognitive function; and the Alzheimer s Disease Functional Assessment and Change Scale (ADFACS), a functional test assessing the patient s ability to perform instrumental activities of daily living (ADLs) and basic ADLs. Safety results were reported for all patients who received open-label treatment. Statistical analysis Efficacy analyses were performed on observed cases up to week 30 of the open-label treatment (54 weeks since the start of the double-blind studies) and summarised using descriptive statistics. Results of outcome measures were presented as mean change from the baseline score. Statistical analyses were conducted using analysis of covariance for the ADAS-cog. All patients randomised into the open-label extension phase of the study were included in the safety analysis. Results Patient disposition Of the 1219 patients enrolled in the two double-blind studies, 885 (72.6%) continued into the open-label phase. In total, 707 (79.9%) of these 885 patients completed the 30-week open-label extension. A subset of 227 (32.1%) patients continued into the additional extension phase, of whom, 194 (85.5%) completed the additional 36 weeks (Figure 1). A higher proportion of patients randomised to the placebo group during the original double-blind phase discontinued during the 30-week open-label phase (23.6%) compared with the two donepezil groups (19.8 and 16.7% of patients from the original 5 and 10 mg/day donepezil groups, respectively). During the 30-week open-label phase, a total of 108 (12.2%) subjects discontinued prematurely due to AEs: 43 (39.8%) originally randomised to placebo, 36 (33.3%) to 5 mg/day donepezil and 29 (26.9%) to 10 mg/day donepezil. During the additional 36-week extension, 19 (8.4%) of the subset of 227 patients withdrew due to AEs. The most common individual AEs resulting in study withdrawal were cerebrovascular accident (4 [1.3%] placebo and 10 [1.7%] donepezil patients), asthenia (four [1.3%] placebo and six [1.0%] donepezil patients) and nausea (five [1.7%] placebo and four [0.7%] donepezil patients). All other AEs that led to study discontinuation occurred in less than 1% of patients. Patient characteristics No significant differences were found in the baseline characteristics of the patients enrolled in this openlabel study (sex, race, mean age, dementia severity) when compared with those of the initial double-blind studies total cohort. Moreover, demographic characteristics were similar regardless of original treatment group (Table 1). Baseline characteristics of the patients enrolled in the double-blind trials have been described previously (Roman et al., 2005). During the 30-week open-label phase, 111 patients (12.5%) remained on the 5 mg/day dose, 78 patients (8.8%) had their dose increased to 10 mg/day but then returned to 5 mg/day and 696 (78.6%) patients escalated to 10 mg/day donepezil and remained at this dose. Nearly all patients (99.9%) took at least one concomitant medication during the study. These included antithrombotic agents (85.4% of the total population); vitamins (42.8%); agents acting on the renin angiotensin system (36.6%); psychoanaleptics, including hypnotics, antidepressants and tranquillisers (38.2%) and serum lipid-reducing agents (31.4%). Efficacy Cognitive function. Results for the primary outcome measure, the ADAS-cog, showed a mean change from double-blind study baseline score to week 54 (24 weeks double-blind plus 30 weeks open-label) of between 0.6 and 1.15 points for patients who had received donepezil at either dose for the entire 54 weeks (Figure 2A). The mean ADAS-cog score for all the donepezil-treated patients who elected to go into the open-label study improved over baseline during the double-blind trials and remained above baseline for the entire open-label phase of the study. This improvement was observed regardless of the initial randomised dosage (5 or 10 mg of donepezil) during the double-blind phase (Figure 2A). The initiation of donepezil treatment in patients who had been randomised to receive placebo in the

308 D. Wilkinson et al. Figure 1 Patient disposition. double-blind trials and who elected to enter the openlabel phase produced no significant improvement in their ADAS-cog scores, but their scores did remain stable, close to their baseline values throughout the additional 30 weeks of treatment (Figure 2A). The MMSE score in both the 5 and 10 mg/day donepezil treatment groups was significantly improved relative to placebo at the end of the double-blind studies and remained improved relative to baseline during the open-label study (Figure 2B). Patients in the delayed start group showed an ameliorated response with a smaller initial improvement, which was maintained above baseline for a shorter time (Figure 2B). Table 1 Patient demographics of the open-label phase Treatment in double-blind phase Placebo n ¼ 301 Donepezil 5 mg/day n ¼ 303 Donepezil 10 mg/day n ¼ 281 Total N ¼ 885 Mean age (SE) 74.6 (0.5) 74.5 (0.5) 74.9 (0.5) 74.7 (0.3) Range 42 92 38 95 47 90 38 95 Male patients (%) 128 (42.5) 125 (41.3) 112 (39.9) 365 (41.2) Mean ADAS-cog score (SE) 19.0 (0.56) 20.1 (0.57) 20.8 (0.61) 20.0 (0.33) Mean MMSE score (SE) 22.1 (0.23) 22.1 (0.22) 21.7 (0.25) 22.0 (0.14)

Long-term use of donepezil for vascular dementia 309 Figure 2 (A) Mean change from baseline ADAS-cog scores during double-blind and open-label treatment. (B) Mean change from baseline MMSE scores during double-blind and open-label treatment. (C) Mean change from baseline CDR-SB scores during double-blind and open-label treatment. (D) Mean change from baseline ADFACS scores during double-blind and open-label treatment. Dosages in figure keys correspond to dosing schedule during double-blind phase only. During the open label extension, all patients received donepezil 5 mg/day for the first 6 weeks and thereafter 10 mg/day, if approved by a clinician (76.6% of patients remained at 10 mg/day through study completion). ol ¼ open label.

310 D. Wilkinson et al. Figure 2 (Continued). Global function. Scores on the CDR-SB, which improved from baseline during the double-blind phase for both the 5 and 10 mg/day donepezil groups, remained improved relative to the double-blind study baseline for 42 weeks of treatment (18 weeks openlabel; Figure 2C). As with the ADAS-cog, patients randomised to the delayed start group did not experience an improvement in their CDR-SB scores following the initiation of donepezil treatment in the open-label phase; however, CDR-SB scores were stable

Long-term use of donepezil for vascular dementia 311 during the 30 weeks of open-label treatment. At the end of the extension phase, there were no notable differences between the treatment groups. Although there was no significant difference between the groups on the ADFACS scale at endpoint, there was a trend suggesting that patients who had been randomised to 10 mg/day donepezil during the double-blind part of the study declined less during the open-label phase compared with either those receiving donepezil 5 mg/day or the delayed start group (Figure 2D). Tolerability Donepezil was well tolerated. Most AEs were mild to moderate in intensity. No new or unexpected AEs attributable to donepezil emerged during long-term treatment. The AEs that occurred during open-label treatment were the same as those observed during the double-blind studies and they occurred with approximately the same frequency in both phases, with 87.6% of patients in the open-label study and 90.2% of patients in the double-blind studies reporting one or more AE. The only AEs that occurred in more than 10% of patients were diarrhoea (12.8%), accidental injury (12.5%) and infection (11.3%) (Table 2). AEs judged to be related to donepezil treatment resulted from expected cholinergic effects, most commonly nausea (occurring in 5.3% of all patients) and diarrhoea (occurring in 8.8% of all patients). All other drug-related treatment-emergent signs and symptoms occurred in less than 5% of the study population. Treatment-emergent AEs related to the cardiovascular system occurred at a similar rate in treated and placebo patients (28.9% vs. 25.6%, respectively). Of the cardiovascular events recorded during this study, hypertension (6.3% of all patients) and stroke (4.7%) were the most common. There were 29 deaths during the study or within 4 weeks of treatment discontinuation: 7 (2.3%) among patients originally randomised to placebo and 22 (3.8%) among patients originally randomised to donepezil in the double-blind studies. The most common causes of death were cerebrovascular accident, myocardial infarction and malignancy. There were four deaths that investigators attributed possibly to study medication: of these, two patients died of a cardiovascular accident, one from sepsis and one from multisystem failure. Serious AEs were reported for 228 patients. The most common were cerebrovascular accident (4.7%) and accidental injury (2.1%). There were no clinically meaningful changes in clinical laboratory values, vital signs, physical examination findings or electrocardiogram status during the study. Discussion Improvements in cognitive function in patients with VaD gained during the 24-week double-blind treatment with donepezil were maintained above baseline during a further 30 weeks of open-label treatment. Patients receiving donepezil (either dose) for 54 weeks had an improvement in their ADAS-cog score of up to approximately one point from their baseline score, suggesting that continuous treatment with donepezil improves and sustains cognitive benefit for at least a year. Interestingly, although patients receiving 10 mg/ day donepezil initially did better than those patients Table 2 Incidence of most commonly occurring TESS (occurring in 5% of patients) during the open-label phase No. (%) of patients experiencing TESS Treatment in double-blind phase a Placebo n ¼ 301 Donepezil 5 or 10 mg/day n ¼ 584 Total N ¼ 885 Body as a whole 132 (43.9) 270 (46.2) 402 (45.4) Accidental injury 36 (12.0) 75 (12.8) 111 (12.5) Infection 29 (9.6) 71 (12.2) 100 (11.3) Pain 28 (9.3) 36 (6.2) 64 (7.2) Nervous system 136 (45.2) 271 (46.4) 407 (46.0) Cardiovascular system 77 (25.6) 169 (28.9) 246 (27.8) Hypertension 14 (4.7) 42 (7.2) 56 (6.3) Digestive system 119 (39.5) 183 (31.3) 302 (34.1) Diarrhoea 47 (15.6) 66 (11.3) 113 (12.8) Nausea 30 (10.0) 29 (5.0) 59 (6.7) Vomiting 18 (6) 26 (4.5) 44 (5.0) TESS ¼ treatment emergent signs and symptoms. a Dosing during double-blind phase only. During the open label extension, all patients received donepezil 5 mg/day for the first 6 weeks and thereafter 10 mg/day, if approved by a clinician (76.6% of patients remained at 10 mg/day through study completion).

312 D. Wilkinson et al. receiving 5 mg/day on this test, after week 34 (i.e. 10 weeks into the open-label phase) patients receiving 5 mg/day donepezil had nonsignificantly better scores on the ADAS-cog that were maintained relative to those of patients receiving 10 mg/day until the end of the study. This pattern has been seen in a previous AD open-label, extension study, where the crossover occurred after about 110 weeks of uninterrupted donepezil therapy (Doody et al., 2001a). The small number of patients in the 5 mg/day group might provide some explanation; however, the reason for this result in the ADAS-cog remains unclear. Both doses of donepezil (5 and 10 mg/day) were beneficial, and in the CDR-SB and the ADFACS tests in this study, the 10 mg/day donepezil dose tended to be the more beneficial dose, suggesting that the discrepant results for the ADAS-cog may be attributable to the variability of the test. By week 42, the mean MMSE score for all patients was one point higher than at baseline. This is in contrast to the decline expected over this period with no treatment. As change in the MMSE score can be directly correlated with clinical measures of function, such as dressing and toileting activities (Galasko et al., 1995), this sustained improvement in the MMSE may represent retention of function, which is important for both patient quality of life (Andersen et al., 2004; Wlodarczyk et al., 2004) and for caregivers. Unlike the ADAS-cog results, MMSE scores attained by patients in the delayed-start group at 42 and 54 weeks were comparable with those of patients who had received donepezil from study start. The ADAS-cog results, however, indicated that there was still a difference in the level of cognitive function between those patients taking donepezil from study start and those with a delay to initiation of active treatment. This may indicate that the ADAS-cog, which was designed specifically to test cognitive deficits associated with AD, is a less sensitive instrument in patients with VaD than the MMSE, which is a general test of cognition. Results with the CDR-SB were similar to those seen with ADAS-cog: that is, improvements gained by the 5 and 10 mg/day donepezil groups during double-blind treatment were mostly maintained above baseline during the open-label phase, but patients who had been delayed in receiving donepezil generally did not improve upon initiation of donepezil treatment. The effects of treatment on function (ADFACS) were mixed. Over the course of the study, the patients ability to carry out ADLs, as measured by this test, generally deteriorated, despite continued donepezil treatment. The evaluation of the ability to perform ADLs in patients with VaD is complicated and can be affected not only by cognition, but also by physical impairments/limitations the patient might have due to stroke, congestive heart failure or to having experienced a heart attack. Nonetheless, once started on therapy, functional ADLs appeared to stabilise in delayed-start patients. The original two double-blind studies only enrolled patients with stable comorbid conditions, often several months (at least 3) after a cardiovascular event. Patients with VaD generally stabilise until the next event, which possibly explains why the original studies showed that efficacy measure scores remained close to baseline among placebo patients (Pratt, 2002), a plateau effect also seen in other VaD clinical trial populations (Kittner et al., 2000). Over time, however, as individual patients experience additional vascular events, accompanied by clinical deterioration either stepwise or, more commonly, gradually this would begin to appear as a progressive group decline. Thus, a positive outcome for this study would include not only a positive result for the treatment group on one or several of the assessment measures, but also stability or less than expected decline. Donepezil was generally well tolerated in this study, in spite of patients having multiple comorbidities and concomitant drug use. Moreover, tolerability data from the 24-week trials were similar to the year-long data, and the extension phase to 90 weeks, albeit with fewer patients. This mirrors what has been observed in clinical studies of AD patients, a population in which safety data are available for up to 4.9 years (Doody et al., 2001b). It is, therefore, not unreasonable to assume that donepezil can be tolerated over the long term in patients with VaD. There are limitations to the present study. The openlabel design of this extension phase does not allow comparison of treated with untreated patients in the longer term. Although not ideal, open-label extension studies do provide important information and are necessary due to the ethical implications of performing long-term, placebo-controlled studies in patients with dementia. Another possible limitation of the study is related to the donepezil dosing schedules. During the double-blind phase, donepezil-treated patients were randomised to receive either the 5 or 10 mg/day dose. On enrolment in the open-label phase, all patients received the 5 mg/day dose for 6 weeks, with some subsequently remaining on this dose, some increasing to the 10 mg/day dose temporarily, and most increasing to the 10 mg/day dose permanently. Although, it is important to consider the possible implications of these different dosage regimens when interpreting the current results, outcomes from the double-blind phase

Long-term use of donepezil for vascular dementia 313 Key points Donepezil was well tolerated over 6 months in patients with VaD; 14.4% discontinued due to AEs, with diarrhoea being the most common (incidence of 8.8%). Patients treated with donepezil for 54 continuous weeks (double-blind study plus extension study) had ADAS-cog scores at endpoint that were higher than at baseline, including those who were randomised to donepezil 5 mg/day in the doubleblind trial. Patients who initiated donepezil during the openlabel extension study exhibited no change in ADAS-cog scores. During the extension study no new safety or tolerability issues were observed. do suggest that there is no significant differential effect between the 5 and 10 mg/day doses. An additional weakness is that the study was designed at a time before the value of executive function testing in VaD patients was understood. Therefore, the outcome measures were designed with the cognitive and functional pathology of patients with AD, rather than VaD, in mind. Although there are similarities and overlap in the cognitive dysfunction associated with the two diseases, there are also areas of difference. Patients with VaD, for example, typically demonstrate disproportionate impairment in executive function, particularly visual memory and attention (Cannata et al., 2002). A reasonable set of dysfunctional domains in VaD was properly assessed using a range of tests to measure different aspects of cognition and function given the time when this study and the preceding double-blind studies were initiated. In conclusion, although the lack of a placebo control group prevents absolute statements regarding the longterm efficacy of donepezil in a VaD patient population, the observed trends suggest that it improves cognition and stabilises function in these patients and that improvements are sustained over the long term. Also, patients treated earlier tended to do better in some measures of cognition and global function than those with a delayed start to their treatment. Thus, despite divergent opinions (Kavirajan and Schneider, 2007), donepezil appears to offer a therapeutic alternative for patients with VaD. Acknowledgements This study was sponsored by Eisai Inc. and Pfizer Inc. Study design, data collection and interpretation of the results were conceived by the authors. Editorial assistance was provided by R. Daniel, PhD, at PPSI and was funded by Eisai Inc. and Pfizer Inc. References Andersen CK, Wittrup-Jensen KU, Lolk A, Andersen K, Kragh-Sorensen P. 2004. Ability to perform activities of daily living is the main factor affecting quality of life in patients with dementia. Health Qual Life Outcomes 2: 52. Black S, Román GC, Geldmacher DS, et al. Donepezil 307 Vascular Dementia Study Group. 2003. Efficacy and tolerability of donepezil in vascular dementia: positive results of a 24-week, multicenter, international, randomized, placebo-controlled clinical trial. Stroke 34: 2323 2330. Cannata AP, Alberoni M, Franceschi M, Mariani C. 2002. Frontal impairment in subcortical ischemic vascular dementia in comparison to Alzheimer s disease. Dement Geriatr Cogn Disord 13: 101 111. Doody RS, Geldmacher DS, Gordon B, Perdomo CA, Pratt RD. 2001a. Open-label, multicenter, phase 3 extension study of the safety and efficacy of donepezil in patients with Alzheimer disease. Arch Neurol 58: 427 433. Doody RS, Stevens JC, Beck C, et al. 2001b. Practice parameter: management of dementia (an evidence-based review). Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 56: 1154 1166. Dubois MF, Hebert R. 2001. The incidence of vascular dementia in Canada: a comparison with Europe and East Asia. Neuroepidemiology 20: 179 187. Galasko D, Edland SD, Morris JC, Clark C, Mohs R, Koss E. 1995. The consortium to establish a registry for Alzheimer s disease (CERAD). part XI. Clinical milestones in patients with Alzheimer s disease followed over 3 years. Neurology 45: 1451 1455. Kavirajan H, Schneider LS. 2007. Efficacy and adverse effects of cholinesterase inhibitors and memantine in vascular dementia: a meta analysis of randomised controlled trials. Lancet Neurol 6: 782 792. Kittner B, De Deyn PP, Erkinjuntti T. 2000. Investigating the natural course and treatment of vascular dementia and Alzheimer s disease. Parallel study populations in two randomized, placebo-controlled trials. Ann NY Acad Sci 903: 535 541. Pratt RD. 2002. Patient populations in clinical trials of the efficacy and tolerability of donepezil in patients with vascular dementia. J Neurol Sci 203 204: 57 65. Roman GC, Wilkinson DG, Doody RS, Black SE, Salloway SP, Schindler RJ. 2005. Donepezil in vascular dementia: combined analysis of two largescale clinical trials. Dement Geriatr Cogn Disord 20: 338 344. Roman G, Black S, Royall DR, et al. 2006. Efficacy and safety of donepezil in vascular dementia: results from the largest double-blind trial in vascular dementia patients. Eur Neuropsychopharm 16 (Suppl 4): S481 S482. Tohgi H, Abe T, Kimura M, Saheki M, Takahashi S. 1996. Cerebrospinal fluid acetylcholine and choline in vascular dementia of Binswanger and multiple small infarct types as compared with Alzheimer-type dementia. J Neural Transm 103: 1211 1220. Wilkinson D, Doody R, Helme R, et al. Donepezil 308 Study Group. 2003. Donepezil in vascular dementia: a randomized, placebo-controlled study. Neurology 61: 479 486. Wlodarczyk JH, Brodaty H, Hawthorne G. 2004. The relationship between quality of life, mini-mental state examination, and the instrumental activities of daily living in patients with Alzheimer s disease. Arch Gerontol Geriatr 39: 25 33.