Investigation of the short- and long-term effects of donepezil on cognitive function in Alzheimer's disease

Authors


Jun Inoue, MA, Department of Psychiatry and Neurology, Hamamatsu University School of Medicine, 1-20-1 Handayama, Hamamatsu, Shizuoka 431–3192, Japan. Email: juninoue@hama-med.ac.jp

Abstract

Background:  Donepezil is effective in maintaining cognitive function in patients with mild to moderate Alzheimer's disease (AD). However, not all patients respond to donepezil. In the present study, we examined the short- and long-term effects of donepezil on cognitive function after 2 years treatment.

Methods:  The present retrospective study was performed on 122 AD outpatients who had been taking donepezil for ≥1 year. All subjects underwent periodic examination of cognitive function (Mini-Mental State Examination (MMSE), Rorschach Cognitive Index (RCI)) and clinical evaluation on the Clinical Dementia Rating (CDR); first before starting treatment and then at 4-month intervals after initiation of donepezil. Patients were divided into three groups: (i) MG, the ‘maintained’ group, in which the global CDR score was maintained during treatment; (ii) DeG, the ‘declined’ group, in which the global CDR score increased; and (iii) DrG, the drop out group, who discontinued the treatment. Changes in scores on the MMSE and RCI were compared before treatment and at 4-month intervals within each group. In addition, over each 4-month period, MMSE and RCI scores were compared between the three groups. Furthermore, to investigate the condition of effective treatment, the reasons why donepezil treatment was discontinued in the DrG were examined.

Results:  The most frequent reason for discontinuation was the appearance of behavioral and psychiatric symptoms of dementia (BPSD), which were more frequently observed in the DrG. Although depression and/or disinclination were more frequent in the MG, hyperactivity was more frequent in the DeG and DrG. Before treatment, patients in the DrG exhibited significantly lower scores on the MMSE than did patients in the MG and DeG. Upon examination 4 months after starting treatment, patients in the MG showed cognitive improvement on the MMSE and RCI, whereas those in the DeG showed cognitive deterioration on the MMSE.

Conclusion:  The results of the present study suggest that when a short-term beneficial effect of the donepezil on cognitive functions is seen, long-term effect may also be expected at 2 years. Periodic clinical evaluation and examination of cognitive function is indispensable for effective donepezil treatment.

INTRODUCTION

Alzheimer's disease (AD) is a neurodegenerative disorder characterized by a progressive course and development in middle-to-late life. This disorder affects almost 10% of individuals who survive beyond the age of 65 years. Donepezil is effective in maintaining cognitive function in mild to moderate AD.1–7 However, not all patients respond to donepezil and prediction of efficacy is difficult before treatment starts. In a previous study, we reported that beneficial short-term effects of donepezil treatment on cognitive functions may predict the long-term effect after 1 year.8 In the present study, we examined the short- and long-term effects of donepezil on cognitive function over 2 years after initiation of treatment. In addition, to investigate the condition of effective treatment, we examined the reasons why treatment was discontinued.

METHODS

In the present study, the diagnosis of dementia and likely prognoses were explained to all subjects and their caregivers, as were the benefits and potential drawbacks associated with donepezil treatment. Informed consent was obtained from all subjects and their caregivers after an oral explanation of the study had been given. This study was approved by the local ethics committee of the mental clinic.

The present study consisted of a retrospective investigation of 122 AD outpatients (28 men, 94 women) who had been taking donepezil for ≥1 year. All patients had visited Okamoto Clinic between January 2000 and April 2006 and fulfillled the diagnostic criteria of probable AD on the National Institute of Neurological and Communicative Disorder and Stroke, Alzheimer's disease and Related Disorders Association (NINCDS-ADRDA).9 None of them had a history of cerebrovascular disease or neurological disease. The diagnosis of AD was supported by data from brain computed tomography (CT) and/or magnetic resonance imaging (MRI). All patients underwent periodic examination of cognitive function with the Mini-Mental State Examination (MMSE)10 and Rorschach test (R-T), as well as clinical evaluation on the Clinical Dementia Rating (CDR),11 before starting donepezil and at 4-monthly intervals thereafter. Episodic behavioral and psychological symptoms of dementia (BPSD) were observed in 18 of the 122 before treatment started (delirium n = 6; delusion n = 3; depression and/or disinclination n = 3; insomnia n = 3; hostility n = 2; anxiety n = 1). Only two the 18 patients were receiving psychiatric medication (clotiazepam) for insomnia before the start of donepezil treatment and they continued to receive their medication until the end of donepezil treatment. Most subjects were not receiving additional psychiatric medication at the start of donepezil treatment.

Independent of the diagnostic assessments, an experienced clinical psychologist performed the MMSE and R-T. The R-T is often used in clinical practice as a psychological examination and consists of a standard set of 10 inkblots; the patient's thinking and association patterns are brought clearly into focus because the ambiguity of the stimulus provides relatively few clues about what are conventional, standard, or normal responses. It has been reported that elderly people suspected of having dementia respond to Rorschach inkblots with unusual and abnormal patterns.12 Our previous study confirmed these findings and, moreover, we found that in patients with moderate AD, unusual or abnormal responses to R-T become more prominent at the 1-year follow up.13 Thus, we developed a cognitive score (Rorschach Cognitive Index (RCI)) that can be used as an indicator of abstract thinking or recognition.14 On the RCI, patient responses to the R-T cards were scored as −1, 0, or +1 and the sum of the scores for 10 cards comprise the RCI. Thus, the RCI score ranges from −10 to 10, and a score of ≤1 suggests a deficit of abstract thinking or recognition in the aged person.

The clinical efficacy of donepezil was determined on the basis of changes in the global CDR score from baseline at 24 months. All patients were divided into three groups: (i) MG, the ‘maintained’ group, in which the global CDR score was maintained during treatment over the period 4–24 months; (ii) DeG, the ‘declined’ group, in which the global CDR score increased; and (iii) DrG, the drop out group, who discontinued the treatment. In the DrG, we examined the reasons why donepezil treatment was discontinued. In the present study, we did not used the CDR Sum of Boxes, but rather the global CDR score because we divided patients into groups (MG and DeG) based on changes in dementia severity at 24 months compared with baseline.

Statistical analysis

In each group, changes in MMSE and RCI scores were calculated and compared before treatment and after each 4, 8, 12, 16, 20 and 24 months of treatment using paired t-test. In addition, during each 4-month period, MMSE and RCI scores were compared among the three groups using Dunnett's multiple comparison test. Only at 24 months were MMSE and RCI scores compared between MG and DeG using Student's t-test. P < 0.05 was considered significant. Statistical analyses were conducted using spss for Windows (version 14.0; SPSS, Chicago, IL, USA).

RESULTS

Patient demographics

Of the 122 patients enrolled in the present study, 87 (18 men, 69 women) were treated with donepezil over a 2-year period and, during that time, underwent periodic evaluation of cognitive function and CDR. Patient demographics for each group are given in Table 1. No significant differences were found between the three groups in terms of age, sex, CDR and RCI before treatment. In the DrG, the MMSE score was significantly lower than that in both the MG (= 0.006) and DeG (= 0.018).

Table 1.  Patient demographics
 MGDeGDrG
  1. Where appropriate, data are the mean ± SD. *P < 0.01 compared with the MG; P < 0.05 as compared with the DeG.

  2. MG, maintained group; DeG, declined group; DrG, drop out group; CDR, Clinical Dementia Rating; MMSE, Mini-Mental State Examination; RCI, Rorschach Cognitive Index.

Sample size652235
Age (years)77.3 ± 7.175.7 ± 5.277.0 ± 5.4
Sex (M/F)13/524/1810/25
CDR score before treatment1.2 ± 0.41.1 ± 0.51.3 ± 0.5
MMSE score before treatment19.5 ± 3.319.8 ± 4.217.2 ± 3.9*
RCI score before treatment−2.2 ± 2.3−1.7 ± 3.1−2.4 ± 3.0

Reasons for discontinuation of donepezil

First, we examine the reasons why donepezil treatment was discontinued. Thirty-five patients (10 men, 25 women) dropped out between 13 and 24 months after starting donepezil treatment. In the DrG, 15 of 35 patients (43%) discontinued treatment in less than 16 months of treatment, 10 of 35 patients (29%) discontinued treatment in less than 20 months, and 10 of 35 patients (29%) discontinued treatment in less than 24 months. At the time when treatment was discontinued, 19 of 35 patients (54%) showed a deterioration on the CDR compared with prior to starting treatment; in 16 of 35 patients (46%), the CDR was maintained at the time of discontinuation of donepezil compared with before treatment.

The reasons for discontinuation of donepezil treatment are given in Table 2. The most frequent reason for discontinuation was the appearance of the BPSD (hyperactivity in seven, hostility in one, insomnia in two, and depression one patient). Ten of these patients received additional psychiatric meditation for the symptoms of the BPSD (tiapride n = 7; zopiclone n = 1; zolpidem n = 1; paroxetine n = 1); BPSD requiring psychiatric medication was observed more frequently in subjects who showed a deterioration on the CDR (7/19) than in subjects in whom the CDR was maintained (3/16). In the DrG, 11 patients continuously exhibited symptoms of BPSD and donepezil treatment was discontinued because of this. Another 10 subjects exhibited episodic symptoms of BPSD (hyperactivity in three, hostility in three, delusion in three, and delirium in one patient), but donepezil treatment was not discontinued because of BPSD in these patients. Other frequent reasons for discontinuation were problems with drug compliance and caregivers' problems. Problems with drug compliance were observed equally in subjects who showed a deterioration on the CDR (4/19) and in those in whom the CDR was maintained (3/16). Caregivers' problems included low motivation for treatment, as well as a caregiver's physical illness or death. Ineffective treatment may contribute to the low motivation for donepezil therapy. Even when treatment was effective, continuation of treatment became difficult in the case of a caregiver's physical illness or death.

Table 2.  Reasons why donepezil treatment was discontinued in the drop out group
 Duration of treatmentTotal
12 months16 months20 months
  1. BPSD, behavioral and psychiatric symptoms of dementia.

BPSD42511
Caregivers' problems4307
Problems with drug compliance2507
Admission to a home for the aged3025
Admission to hospital due to physical illness1023
Patient death1012

During the treatment period, three subjects were hospitalized due to physical illness and dropped out of the treatment. One male patient was hospitalized at 77 years of age (after 12 months treatment) because of prostate cancer (diagnosed and treated before donepezil treatment), one female patient was hospitalized at 73 years of age (after 21 months treatment) because of cerebral hemorrhage, and one male patient was hospitalized at 80 years of age (after 22 months treatment) because of pneumonia. During the treatment period, two female pateints died at 80 and 85 years of age (after 13 and 22 months treatment, respectively). Although these patients' families advised that the cause of death was acute heart failure, autopsies were not performed and detailed information is not available.

Transition in cognitive function in the MG, DeG, and DrG

Over the 2-year treatment period, 22 of 87 subjects (25%) exhibited deterioration on the CDR compared with before treatment (DeG), whereas the CDR was maintained in 65 of 87 subjects (75%) over 4–24 months of donepezil treatment (MG). Transitions on the CDR, MMSE, and RCI in the MG, DeG, and DrG are detailed in Tables 3–5.

Table 3.  Transition on the Clinical Dementia Rating in the maintained group, declined group, and drop out group
GroupBefore treatmentDuration of treatment
4 months8 months12 months16 months20 months24 months
  1. Data show the mean ± SD.

  2. MG, maintained group; DeG, declined group; DrG, drop out group.

MG1.2 ± 0.41.1 ± 0.41.1 ± 0.41.1 ± 0.41.1 ± 0.51.1 ± 0.51.2 ± 0.4
DeG1.1 ± 0.51.1 ± 0.51.1 ± 0.51.3 ± 0.71.6 ± 0.61.8 ± 0.82.1 ± 0.8
DrG1.3 ± 0.51.3 ± 0.51.4 ± 0.61.7 ± 0.71.9 ± 0.82.0 ± 1.0 
Table 4.  Transition on the Mini-Mental State Examination in the maintained group, declined group, and drop out group
GroupBefore treatmentDuration of treatment
4 months8 months12 months16 months20 months24 months
  1. Data show the mean ± SD. *P < 0.01, **P < 0.05 compared with before treatment within the same group.

  2. MG, maintained group; DeG, declined group; DrG, drop out group.

MG19.5 ± 3.320.3 ± 3.0**20.4 ± 3.4**20.1 ± 3.420.2 ± 3.919.3 ± 3.719.7 ± 3.6
DeG19.8 ± 4.218.3 ± 4.1**18.1 ± 4.2*17.4 ± 4.3*16.3 ± 4.2*15.7 ± 4.7*13.9 ± 4.5*
DrG17.2 ± 3.918.5 ± 3.7**17.8 ± 3.916.1 ± 4.914.0 ± 6.2*13.4 ± 7.3 
Table 5.  Transition on the Rorschach Cognitive Index in the maintained group, declined group, and drop out group
GroupBefore treatmentDuration of treatment
4 months8 months12 months16 months20 months24 months
  1. Data show mean ± SD. *P < 0.01, **P < 0.05 compared with before treatment within the same group.

  2. MG, maintained group; DeG, declined group; DrG, drop out group.

MG−2.2 ± 2.3−1.2 ± 2.3*−1.1 ± 3.0*−1.4 ± 2.7**−1.4 ± 2.7**−1.5 ± 2.6**−1.4 ± 2.7**
DeG−1.7 ± 3.1−2.0 ± 2.8−2.1 ± 3.0−1.6 ± 3.5−2.4 ± 3.1−3.0 ± 3.0−3.7 ± 3.5**
DrG−2.4 ± 3.0−1.7 ± 3.2−2.5 ± 2.9−3.5 ± 3.8−3.7 ± 3.6−4.2 ± 4.6 

Of the 22 subjects in the DeG, deterioration on the CDR was observed in one patient at 4 months, one patient at 8 months, five patients at 12 months, six patients at 16 months, four patients at 20 months, and five patients at 24 months.

At the evaluation performed after 4 months treatment, significantly higher MMSE (t = 2.46; P = 0.017) and RCI (t = 3.74; P < 0.001) scores were seen in the MG compared with before treatment. Conversely, a significantly lower MMSE score was seen in the DeG compared with before treatment (t = 2.22; P = 0.038). Although the MMSE score in the DrG was significantly higher than before treatment (t = 2.74; P = 0.01), the MMSE score in this group remained significantly lower than that in the MG (= 0.012).

After 8 months treatment, the MMSE score in the MG was significantly higher than before treatment (t = 2.51; P = 0.015). However, in the DeG, the MMSE score at 8 months (t = 3.16; P = 0.005), 12 months (t = 3.67; P = 0.001), 16 months (t = 5.82; P < 0.001), 20 months (t = 6.38; P < 0.001), and 24 months (t = 9.17; P < 0.001) treatment was significantly lower than before treatment. The RCI score was significantly higher in the MG after 8 months (t = 3.78; P < 0.001), 12 months (t = 2.52; P = 0.014), 16 months (t = 2.19; P = 0.032), 20 months (t = 2.19; P = 0.032), and 24 months (t = 2.27; P = 0.026) treatment than before treatment. The RCI score in the DeG at 24 months was significantly lower than before treatment (t = 2.67; P = 0.014).

Comparing the MG and DeG, there was a higher MMSE score in the MG at 12 months (P = 0.038), 16 months (P = 0.002), 20 months (P = 0.009), and 24 months (t = 5.50, P < 0.001). In addition, the RCI score was significantly higher in the MG than the DeG at 24 months (t = 2.74; P = 0.01). Comparison of the MG and DrG revealed that the MMSE score was significantly higher in the MG than the DrG at 4 months (P = 0.044), 8 months (P = 0.006), 12 months (< 0.001), and 16 months (P = 0.001). In addition, the RCI score was significantly higher in the MG than the DrG at 12 months (P = 0.02) and 16 months (P = 0.036).

Comparison of BPSD in the MG, DeG, and DrG

The most frequent reason for discontinuation was the appearance of BPSD, so this parameter was compared between the MG, DeG, and DrG. Before treatment, BPSD (depression and/or disinclination n = 3; delusion n = 3; insomnia n = 3; delirium n = 3; hostility n = 1; anxiety n = 1) was observed 14 of 87 patients (16%) who completed 2 years treatment; two of 14 patients received psychiatric medication (clotiazepam) for insomnia before starting donepezil treatment and continued to receive this medication until the end of donepezil treatment. The remaining 12 of 14 subjects did not receive any psychiatric medication and BPSD disappeared during the early stages of the treatment. Conversely, BPSD (delirium n = 3; hostility n = 1) was observed in four of 35 (11%) subjects in the DrG before the start of donepezil treatment. In these four patient, BPSD disappeared during the early stages of treatment without the need for additional psychiatric medication.

Of the 87 subjects who received donepezil treatment over 2 years, five (6%) received additional medication for symptoms in the digestive system from the early stages of donepezil treatment. Furthermore, nine of 87 patients (10%) who completed the 2 years treatment exhibited continuous BPSD during the treatment period. As noted earlier, 14 of 87 (16%) subjects showed BPSD before treatment and although only two patients received additional psychiatric medication, BPSD disappeared during the early stages of donepezil treatment without the need for additional psychiatric medication in most cases. Only two of 14 patients exhibited another BPSD during the 2-year treatment period, and 12 of 14 subjects did not exhibit BPSD during the treatment period. In the MG, three subjects (5%) showed continuous BPSD (depression and/or disinclination n = 3), and two (3%) showed episodic BPSD (hyperactivity in one and hostility in another). Only three subjects in the MG (5%) received additional psychiatric meditation (paroxetine in all cases). In the DeG, four patients (18%) showed continuous BPSD (hyperactivity in three and insomnia in one), and one (5%) showed episodic BPSD (hyperactivity). In the DeG, four of 22 patients (18%) received additional psychiatric meditation (tiapride n = 3; flunitrazepam n = 1). In the DrG, 21 of 35 subjects (60%) showed continuous or episodic BPSD, and 10 patients (29%) received additional psychiatric meditation. Thus, BPSD was more frequently observed in the DrG. Although, depression and/or disinclination were more frequent in the MG, hyperactivity was more frequent in the DeG and DrG.

DISCUSSION

A recent community-based study indicated that the efficacy of donepezil treatment was questionable.15 In our previous study, 81 of 109 outpatients (74%) who received 1 year donepezil treatment maintained cognitive function. In addition, in the present study, 65 of 87 outpatients (75%) who received donepezil treatment over 2 years maintained cognitive function and improved abstract thinking or recognition after 2 years follow up. These results suggest that donepezil is effective in the maintenance of cognitive function in mild to moderate AD not just after 1 year follow up, but also after 2 years. However, 35 subjects (29%) discontinued treatment after 13–24 months and cognitive deterioration was observed in more than half (54%) of those who dropped out. The MMSE score before treatment in the DrG was significantly lower than not only than that in the MG, but also the DeG. Although half the subjects in the DrG did not show any deterioration on the CDR at the last examination, there was marked cognitive deterioration on the MMSE and the RCI in the DrG. These data suggest that many subjects in the DrG had a more severe cognitive deficit and are therefore likely to show cognitive deterioration at an early stage.

Furthermore, 60% of subjects in the DrG showed continuous or episodic BPSD, and 29% received additional psychiatric meditation. This is compared with 23% of patients in the DeG who showed BPSD, 18% of whom received additional psychiatric meditation because of the BPSD. In the MG, only 8% of subjects showed BPSD, with 5% receiving additional psychiatric meditation because of the BPSD. Conversely, subjects in the MG rarely showed hyperactivity or hostility; the BPSD in the MG consisted mainly of depression and/or disinclination. Thus, cognitive deterioration may cause BPSD such as hyperactivity or hostility. These BPSD may increase the physical and psychological burden of the caregivers, and may consequently decrease the motivation of caregivers to continue with donepezil treatment.

There were no significant differences in age, CDR, MMSE, and RCI between the MG and DeG before treatment. These data suggest that prediction of treatment efficacy for donepezil is difficult before treatment. Conversely, at the 4-month examination, patients in the MG showed cognitive improvement on the MMSE and RCI, whereas those in the DeG showed cognitive deterioration on the MMSE. Furthermore, improvements on the RCI in the MG were not a short-term effect, but rather a long-term effect. Although, the neurological and pharmacological mechanisms responsible for the improvement on the RCI are not clear, the effect of donepezil on attention may account for the improvement on the RCI.

In a previous study, we reported that the short-term effect on cognitive function may predict the long-term effect after 1 year donepezil treatment. The results of the present study suggest that when a beneficial short-term effect of donepezil on cognitive functions is seen, a long-term effect may also be expected after 2 years. In the DeG, donepezil treatment was not necessarily ineffective in terms of activities of daily living and quality of life of both patients and caregivers because they hoped for the continuance of treatment. However, the inhibition by donepezil of the progression of cognitive decline in the DeG was obviously inferior to that in the MG. When a short-term beneficial effect is not seen, cognitive function may decrease further in the future. We can give patients and their caregivers some feedback regarding the change after cognitive decline. We can explain the appearance of BPSD and difficulties of daily life as a result of the decline in cognitive functions to a subject's caregivers and assist them to establish and/or enhance the nursing situation. Thus, periodic clinical evaluation and examination of cognitive function is indispensable for effective donepezil treatment.

Over the 2-year treatment period, three subjects (one woman, two men) were hospitalized because of a physical illness and two female patients died. Detailed information was not available in any case, so it is not known whether there is a causal relationship between donepezil treatment and the physical illness or death. Further close examination is required for adverse effects of donepezil in the long term.

CONCLUSION

When a beneficial short-term effect of donepezil on cognitive functions was seen, a long-term effect could also be expected. Periodic clinical evaluation and examination of cognitive function is indispensable for effective donepezil treatment. In particular, the RCI is a useful cognitive predictor of donepezil treatment. When a beneficial short-term effect on cognitive functions is not seen during the early stages of donepezil treatment, it is necessary to consider other treatment strategies or to prepare to lighten the burden imposed on caregivers.

Ancillary