Efficacy and Safety of Sustained Release Donepezil High Dose versus Immediate Release Donepezil Standard Dose in Japanese Patients with Severe Alzheimer’s Disease: A Randomized,Double-Blind Trial

Abstract

Background: Donepezil is an established treatment for mild, moderate, and severe Alzheimer’s disease (AD). An international study demonstrated superior efficacy of sustained release (SR) 23 mg/day donepezil over immediate release (IR) 10 mg/day donepezil for cognitive function, but not global function in moderate-to-severe AD.

Objective: To demonstrate the superiority of SR 23 mg/day donepezil over IR 10 mg/day donepezil in Japanese patients with severe AD (SAD).

Methods: In this multicenter, randomized, double-blind, parallel-group study, Japanese outpatients with SAD were randomly assigned to continue IR 10 mg/day or switch to SR 23 mg/day for 24 weeks. Endpoints included the Severe Impairment Battery (SIB), Clinician’s Interview-Based Impression of Change Plus Caregiver Input (CIBIC-plus), and safety.

Results: Overall, 166 and 185 patients were randomized to receive IR 10 mg/day and SR 23 mg/day, respectively. SR 23 mg/day was not statistically superior to IR 10 mg/day by SIB (least squares mean difference [LSMD]: 0.0; 95% confidence interval [CI]: –1.7, 1.8; p = 0.981) or CIBIC-plus (LSMD: 0.2; 95% CI: 0.0, 0.4; p = 0.080). Common adverse events in the SR 23 mg group were decreased appetite, vomiting, diarrhea, and contusion. Safety findings were consistent with known safety profiles of donepezil.

Conclusion: SR 23 mg/day donepezil was not superior to IR 10 mg/day donepezil regarding the efficacy endpoints for Japanese SAD. Considering that a 10 mg/day dose is approved for SAD in Japan, the present findings suggest that IR 10 mg/day donepezil is the optimal dosage for Japanese patients with SAD.

Keywords

Alzheimer’s disease cholinesterase inhibitors donepezil double-blind study Japan

INTRODUCTION

Currently, four standard drugs, donepezil, galantamine, rivastigmine, and memantine, are used globally to treat Alzheimer’s disease (AD) symptoms.Therefore, for the successful treatment of AD, it is important to tailor the use of these four drugs according to the patient’s symptoms to maintain their daily living functions for as long as possible. Unfortunately, patients with more advanced AD have fewer therapeutic options. Only donepezil and memantine are indicated for severe AD (SAD), and a combination of these drugs is often necessary. However, the potential benefit of this combination therapy is controversial [1, 2]. The maximum approved dose of donepezil in Japan is 10 mg/day, based on the results of an earlier study [3].

It was reported that acetylcholinesterase (AChE) activity, measured by positron emission tomography (PET), in the cerebral cortex of patients with mild-to-moderate AD, was decreased by approximately 30% in those receiving 10 mg/day donepezil, as compared with AChE activity before drug administration [4, 5]. Therefore, AChE activity in the cerebral cortex can be reduced by administering higher doses of donepezil, even in patients treated with 10 mg/day donepezil [6]. These findings suggest symptoms of AD might be improved and maintained by further activating intracerebral cholinergic neurons through an increased dose of donepezil.

A high dose of 23 mg/day donepezil was confirmed to be more effective than the standard dose of 10 mg/day donepezil in an overseas phase III study (Study E2020-G000-326; hereinafter referred to as Study 326) [7]. Study 326 enrolled patients with moderate-to-severe AD who had been treated with 10 mg/day donepezil for ≥3 months and who were randomized to either continue the same dose or to receive the high dose of 23 mg/day for 24 weeks. The co-primary endpoints were cognitive and global functions, which were evaluated using the Severe Impairment Battery (SIB) and Clinician’s Interview-Based Impression of Change Plus Caregiver Input (CIBIC-plus), respectively. The study revealed a significant difference in SIB total scores, but not CIBIC-plus scores, between the two groups. Subgroup analyses revealed that 23 mg/day donepezil displayed significantly greater efficacy than 10 mg/day donepezil in terms of SIB total scores and CIBIC-plus in the US population and in patients with more advanced AD.

The supplemental new drug application for donepezil 23 mg/day was approved in the US in 2010, although this dose is not yet marketed in Japan.

Based on this background, the present study (hereinafter referred to as Study 343) was conducted at 69 Japanese sites to examine the efficacy and safety of 23 mg/day donepezil in Japanese patients with SAD using donepezil 10 mg/day as the active comparator.

PATIENTS AND METHODS

This study was conducted in accordance with ethical principles based on the Declaration of Helsinki, the standards specified in the Pharmaceutical Affairs Law, and Good Clinical Practice. The study was approved by ethics committees/institutional review boards at each participating institution. This trial was registered on ClinicalTrials.gov (identifier: NCT01539031).

Methods

This study was designed as a multicenter, randomized, active-controlled, double-blind, parallel group comparative study followed by an uncontrolled, open-label extension to compare SR 23 mg/day donepezil with the control drug, immediate-release (IR) 10 mg/day donepezil.

The study consisted of a 4-week screening/baseline period, a 24-week double-blind treatment period, a 4-week double-blind transition period, and a 24-week open-label extension period. Subjects were asked to visit every 6 weeks during the double-blind treatment period. The design and results of the transition and extension periods are not reported here.

Patients

Outpatients with SAD aged ≥50 years who had been continuously treated with 10 mg/day donepezil for ≥3 months before the screening period were eligible if they met all of the inclusion criteria and none of the exclusion criteria. The eligible MMSE score ranges of 1–12 were determined to match a previous study of Japanese patients with SAD [3]. Inclusion and exclusion criteria are listed in the Supplementary Methods.

Study treatments in the 24-week double-blind treatment period

After the baseline examination and immediately before the start of the double-blind treatment period, eligible patients were randomized (1:1 ratio) to either continue IR 10 mg/day (IR 10 mg group) or receive SR 23 mg/day (SR 23 mg group). Randomization was achieved using the investigational product assignment codes (key codes) prepared using uniform random numbers generated with the RANUNI function in the DATA Step in SAS software version 9.3 (SAS Institute, Cary, NC, USA). The variable block design was used with block sizes of 2 and 4.

Patients allocated to the SR 23 mg group were orally administered one IR 5 mg tablet and one IR 10 mg tablet/day (15 mg/day) at Weeks 1–4, followed by one IR 10 mg/day placebo tablet and one SR 23 mg/day tablet at Weeks 5–24. Patients allocated to the IR 10 mg group were orally administered one IR 5 mg placebo tablet and one IR 10 mg/day tablets at Weeks 1–4, followed by one IR 10 mg/day tablet and one SR 23 mg/day placebo tablet at Weeks 5–24. All tablets were administered once daily in the morning.

Endpoints

The co-primary efficacy endpoints were the change in SIB from baseline to Week 24 of the double-blind treatment period with last observation carried forward (LOCF) to assess cognitive function and the CIBIC-plus at Week 24 (with LOCF) to assess global clinical changes.

The SIB is a neuropsychological test used to assess cognitive dysfunction in severely impaired dementia patients, which is performed by interview with the patient [8]. SIB consists of nine subscales used to evaluate nine domains including social interaction, memory, orientation, language, attention, praxis, visuospatial ability, construction, and orienting to name. Scores range from 0 (worst score) to 100 (best score).

The CIBIC-plus is a semi-structured global rating scale derived from a separate interview with a patient and his/her caregiver by an experienced independent rater. It is designed to measure various domains: general, mental/cognitive state, behavior, and activities of daily living. The rater assesses clinically meaningful changes from baseline at each visit using a 7-point scale as follows: 1 = marked improvement; 2 = moderate improvement; 3 = minimal improvement; 4 = no change; 5 = minimal worsening; 6 = moderate worsening; and 7 = marked worsening [9, 10]. The caregivers were interviewed first in the present study.

Secondary efficacy endpoints included SIB values at each visit, the changes in SIB from baseline to each visit (Weeks 6, 12, 18, and 24), and the CIBIC-plus at the time of each assessment.

Safety-related endpoints included adverse events (AEs), laboratory findings (hematology and biochemistry), vital signs (systolic blood pressure, diastolic blood pressure, and pulse rate), body weight, and 12-lead ECG. All ECGs were measured by standardized ECG equipment and were evaluated at a central laboratory in the US (BioClinica, Princeton, NJ, USA) according to a standard procedure.

Pharmacokinetics (PK) of donepezil

Blood samples were collected at each visit from Week 6 of the treatment period, by which time the plasma donepezil concentration was considered to have reached a steady state, through to Week 24 in order to measure plasma donepezil concentrations and determine its PK characteristics. At any visit, blood samples (3 mL) were obtained at the following times: (1) before taking the study drug; (2) 1–3 h; (3) 3–6 h; and (4) 6–8 h after drug administration. Plasma donepezil concentrations were measured using validated liquid chromatography tandem mass spectrometry.

Statistical methods

Sample size calculation

The sample size was determined based on re-sampling simulation using data (SIB and CIBIC-plus) from a subgroup of American patients with Mini-Mental State Examination (MMSE) score ranges of 1–12 in Study 326 [7]. Three-hundred and thirty patients was the planned target sample size, to provide a power of ≥80% to detect statistically significant difference between the groups, at a two-sided significance level of 5% , using a primary model from re-sampled data with co-primary endpoints, with the assumption that approximately 10% of patients were to be excluded from analyses.

Efficacy

Data in the 24-week double-blind period were analyzed for the full analysis set (FAS; the population of patients with ≥1 point of evaluable efficacy data after administration of the investigational product) and the per-protocol set (PPS; the population of patients complying with the protocol until the completion of the assessment at Week 24).

The primary analysis was conducted using the FAS, and sensitivity analyses were conducted using the PPS. With regard to the evaluation visit for primary analysis (Week 24), the LOCF was obtained by handling evaluable final data after the study treatment. To determine changes from the baseline to Week 24 (LOCF) of SIB, the superiority of SR 23 mg/day was tested against IR 10 mg/day by ANCOVA with terms for treatment group (main effect) and baseline value (covariate). Similarly, for CIBIC-plus at Week 24 (LOCF), an ANCOVA model with terms for treatment group (main effect) and baseline value (CIBIS; covariate) were also used to test the treatment effects.

Secondary endpoints were analyzed using the FAS (see Supplementary Methods for details of the analyses of secondary endpoints).

All statistical tests were analyzed using two-tailed tests, where p < 0.05 indicated statistical significance.

Pharmacokinetics

PK evaluations were performed using the PK analysis set (all patients with at least one evaluable plasma concentration). Summary statistics were calculated for the plasma concentrations of donepezil according to dose and sampling time points. The relationships between the plasma concentration of donepezil and efficacy endpoints were also examined.

The steady state plasma concentration predicted by population PK analysis, and the change from baseline for SIB and CIBIC-plus at Week 24 as efficacy endpoints were used for this analysis.

Safety

Safety analyses were conducted using the safety analysis set (all patients with at least one safety variable recorded after starting administration of the study drug in the double-blind period). The incidences of AEs overall, and according to system organ class and preferred MedDRA term, were calculated for the 24-week double-blind treatment period. Serious AEs were collected according to Good Clinical Practice guidelines, and AE severity was assessed using a three-point scale: mild, moderate, and severe.

Laboratory values and vital signs were measured at each visit and changes from baseline were calculated. QTcF data are presented categorically using the following classifications: >450 ms/>480 ms/>500 ms, > Δ30 ms/> Δ60 ms. All statistical analyses were performed using SAS software version 9.3.

RESULTS

Patient disposition

Overall, 351 patients were randomized and treated, with 166 in the IR 10 mg group and 185 in the SR 23 mg group (Fig. 1). Of these, 274 patients completed the 24-week treatment period (141 in the IR 10 mg group and 133 in the SR 23 mg group). Overall, 77 patients discontinued treatment (Fig. 1), mainly because they experienced AEs (13 in the IR 10 mg group and 29 in the SR 23 mg group).

Patient baseline characteristics

Table 1 shows the baseline characteristics of 340 patients in the FAS. Females accounted for 69.4% of the patients, the mean age was 76.0 years, and the mean body weight was 51.9 kg. Overall, 53.8% received concomitant memantine. The baseline MMSE score was 8.7. There were no notable differences in baseline characteristics between the two groups.

Co-primary endpoints (SIB total score and CIBIC-plus)

Table 2 shows the results for the co-primary endpoints. The LS mean±SE change in the SIB total score from baseline to Week 24 (LOCF) was –2.7±0.6 in the IR 10 mg group and –2.7±0.6 in the SR 23 mg group, corresponding to a LS mean difference of 0.0 (95% CI: –1.7, 1.8; p = 0.981). The LS mean±SE CIBIC-plus score at Week 24 (LOCF) was 4.5±0.1 in the IR 10 mg group and 4.7±0.1 in the SR 23 mg group, corresponding to a LS mean difference of 0.2 (95% CI: 0.0, 0.4; p = 0.080). Therefore, SR 23 mg/day was not superior to IR 10 mg/day in either of the co-primary endpoints.

The PPS analyses yielded similar results to the FAS analyses.

Secondary endpoints

Figure 2 shows changes in the SIB total score from baseline to each visit. SIB total scores started to decrease in both groups at Week 6 and continued to decrease until Week 24. These changes did not differ significantly between the two groups.

The LS mean of CIBIC-plus scores exceeded 4 (worsened) from Week 6 to Week 24 in both groups, and was significantly lower (clinically better) in the IR 10 mg group than in the SR 23 mg group at Week 6. However, CIBIC-plus was not significantly different between the two groups at any other visit. Figure 3 shows that the distribution of CIBIC-plus scores at Week 24 (LOCF) was not significantly different between the two groups.

Subgroup analysis

To explore the superiority of SR 23 mg/day to IR 10 mg/day on the change in the SIB total score, prespecified analyses were performed in subgroups of patients according to the concomitant use of memantine, and post hoc analyses were performed for subgroups of patients according to MMSE score at screening (≤median/>median) and the duration of pretreatment with 10 mg/day donepezil before enrollment (≤median/>median).

In the subgroup with lower MMSE scores (i.e., more severe AD), the SIB total score tended to worsen in the IR 10 mg group compared with the SR 23 mg group, although the difference was not statistically significant (Fig. 4a).

The changes in SIB total scores were similar between the treatment groups in both subgroups of patients (duration of pretreatment≤median/>median) (Fig. 4b) as well as subgroups according to the concomitant use of memantine (Fig. 4c).

These results did not indicate that SR 23 mg and IR 10 mg efficacies were notably influenced by the severity of disease, the duration of donepezil pretreatment, or the concomitant use of memantine.

Plasma concentrations of donepezil

The mean plasma concentration of donepezil was highest at 3–6 h after administration in both the SR 23 mg and IR 10 mg groups. It was 2.0–2.3 times higher in the SR 23 mg group than in the IR 10 mg group at all measured points (see Supplementary Results for plasma concentrations of donepezil for details).

Relationship between the plasma concentrations of donepezil and efficacy

The relationships between the plasma concentrations of donepezil with the change in SIB from baseline and CIBIC-plus (PK–efficacy) are shown in Fig. 5. Changes in SIB from baseline and CIBIC-plus were comparable in all five groups of plasma concentrations of donepezil, although there was a slight tendency toward a worsening of both endpoints (i.e., decrease in SIB and increase in CIBIC-plus) with increasing plasma concentrations of donepezil.

Safety

Table 3 shows the AEs, including individual types of AEs with an incidence ≥2% in either group (by preferred term). The total AE rate was 72.4% in the SR 23 mg group (134/185) and 59.0% in the IR 10 mg group (98/166). AEs with an incidence ≥5% in the SR 23 mg group and with an incidence ≥2 fold higher than those in the IR 10 mg group were decreased appetite, vomiting, diarrhea, and contusion. Most AEs were determined to be mild or moderate. The incidence of severe AEs was 6.6% in the IR 10 mg group and 6.5% in the SR 23 mg group.

The incidence of serious AEs was 8.4% in the IR 10 mg group and 12.4% in the SR 23 mg group. Only one AE (subarachnoid hemorrhage) resulted in death, and occurred in one patient in the IR 10 mg group. This event was considered not to be related to the study drug. Serious AEs that only occurred in the SR 23 mg group and that occurred in ≥2 patients included dehydration and urinary tract infection (in two patients each); these events were considered not to be related to the study drug.

The incidence of AEs leading to discontinuation of the test drugs was 9.0% in the IR 10 mg group and 17.8% in the SR 23 mg group. AEs leading to discontinuation that occurred in ≥2 patients in the SR 23 mg group with an incidence ≥2 fold higher than those in the IR 10 mg group were decreased appetite, QT prolonged, nausea, and vomiting. Investigators did not deny the causal relationships of these events with the study drug. All of the events resolved except for decreased appetite in one patient.

There were no notable differences regarding abnormal changes in laboratory values and vital signs except body weight between the treatment groups. Overall, 3.0% of patients in the IR 10 mg group and 19.5% in the SR 23 mg group showed a weight decrease of at least 7% compared with baseline. In one patient in the IR 10 mg group and four in the SR 23 mg group, the decreased body weight was considered to probably/possibly be related to the study drug.

There were no notable differences regarding changes in ECG parameters, except QT/QTcF, between the two groups. A prolongation of QTcF from baseline was observed at each visit after Week 6 only in the SR 23 mg group (mean ΔQTcF range 2.2 to 6.9 ms). A maximum QTcF value of >450 ms was observed in 11.3% of patients in the IR 10 mg group and 20.7% of patients in the SR 23 mg group. A maximum ΔQTcF of >30 ms was seen in 11.9% of IR 10 mg group patients and in 25.1% of SR 23 mg group patients. No IR 10 mg group patients had a maximum QTcF >500 ms, but this was seen in 0.6% of SR 23 mg group patients, and a maximum ΔQTcF of >60 ms was recorded in 2.5% and 3.4% of patients in each group, respectively.

DISCUSSION

This multicenter, randomized, double-blind, parallel group clinical trial to study the superiority of SR 23 mg/day donepezil over IR 10 mg/day donepezil in Japanese patients with SAD demonstrated that SR 23 mg/day donepezil was not superior to IR 10 mg/day donepezil. No difference was found between the dosages in terms of the SIB total score (Table 2) or the CIBIC-plus score (Table 2) in our study (Study 343).

The worsening of the SIB total score of the SR 23 mg group over time was similar to that in the IR 10 mg group (Fig. 2). The distribution of CIBIC scores was favorable to IR 10 mg/day in all seven determinations (Fig. 3).

A previous clinical study with a similar design was conducted in patients with moderate and severe AD (MMSE 0–20; Study 326) [7]. This revealed that 23 mg/day was superior to 10 mg/day in terms of SIB but not CIBIC-plus. However, in subgroup analyses, both SIB and CIBIC-plus were significantly better in the US subgroup and the subgroup with more advanced disease (MMSE 0–16) following 23 mg/day compared with 10 mg/day.

Therefore, the number of patients in the current study was designed based on the observed efficacy of SR 23 mg/day in the subgroup of US patients with SAD (MMSE 1–12) in Study 326. However, the results of our study differed from this assumption. In addition, subgroup analyses of Study 326 suggested a significant benefit of 23 mg/day over 10 mg/day regardless of baseline MMSE (0–9, 10–20), the duration of donepezil pretreatment [13], and the concomitant use of memantine [7, 13]. By contrast, no benefits were suggested in the current study in those subgroups (Fig. 4a–c).

Higher doses of donepezil were tested based on the assumption that additional benefit could be obtained by maintaining higher plasma concentrations of donepezil [7]. As expected, the current study indicates that donepezil plasma concentrations at all time-points in the SR 23 mg group were approximately twice as high as those in the IR 10 mg group. Nevertheless, changes in SIB total score and CIBIC-plus (Fig. 5a, b) suggest that cognitive functions and global clinical changes tended to be maintained or slightly worsened in this study by increasing the plasma concentration of donepezil through increasing the dose to ≥10 mg. Kuhl et al. [11] reported that AChE activity in the AD cerebral cortex was decreased by 26% after the administration of 5 mg/day donepezil for ≥5 weeks compared with activity before treatment. They also reported that the decrease of AChE activity was maintained at 27% compared with that before treatment after a dose increase to 10 mg/day administered for ≥8 weeks. Similarly, it is possible that cerebral cortex AChE activity in our patients did not decrease sufficiently compared with activity before enrollment in patients in the 23 mg/day group.

In Study 326, the PK–efficacy relationship for SIB showed an inverted U-shaped graph, with a peak at the median concentration of 23 mg. Meanwhile, the PK–efficacy relationship for CIBIC-plus showed a U-shaped graph, with a trough at the median concentration between 10 mg/day and 23 mg/day [12]. These results indicate that 23 mg/day provided greater benefit for the efficacy endpoints than 10 mg/day among patients with moderate-to-severe AD. However, further increases in exposure to donepezil, via intrinsic or extrinsic factors, could lead to suboptimal efficacy in addition to potential safety concerns [12]. Thus, it appears that the plasma concentration of donepezil at which maximum efficacy is obtained has an upper limit in both Japanese and non-Japanese patients. It is difficult to explain the different tendencies of the PK–efficacy relationship between Japanese and non-Japanese patients because of the existence of limitations of the present PK–efficacy assessment on the following points: first, the study periods, the presence/absence of dose titration, and disease severity differed between the present study and Study 326; second, the variability of each efficacy parameter in the PK–efficacy plot was relatively large; and third, the PK–efficacy relationship was evaluated post hoc without a sufficient number of patients. Another study limitation might be the lack of appropriateness of the LOCF method to address the scientific question of interest, as there were higher numbers of dropouts in the SR 23 mg group compared with the IR 10 mg group. The current study (Study 343) aimed to verify the hypothesis adopted in a prior overseas study (Study 326) and its sample size was calculated on the basis of the results of Study 326. Because LOCF was used in Study 326, we decided to also use LOCF in this study, although other analysis methods such as mixed model for repeated measures are more frequently being applied assuming the data are missing at random. However, to retain the same methodology as used in Study 326 for comparison, we used LOCF in this study.

As shown in Study 326, a higher AE incidence was observed in the SR 23 mg group than in the IR 10 mg group. It is important to note that all patients in Studies 326 and 343 were treated with IR 10 mg/day for ≥3 months before randomization, which implies an established tolerability for this dose of donepezil.

The most common AEs in our study were decreased appetite and gastrointestinal AEs, which are related to cholinesterase inhibition, and are consistent with Study 326 findings.

The incidence of decreased appetite in this study was 9.7% (18/185) and 4.2% (7/166) in the SR 23 mg group and IR 10 mg group, respectively, and was higher than that in Study 326 (5.3% [51/963] and 1.7% [8/471], respectively) [7]. However, the incidence of decreased appetite in our study was comparable to that in an Asian population (10.3% [15/145] and 2.6% [2/78], respectively) [14].

In Study 326, more patients in the 23 mg/day group who weighed <55 kg reported nausea, vomiting, and decreased weight than patients weighing ≥55 kg. This finding may be related to the higher plasma exposure of donepezil in patients with a lower body weight (Aricept Package,US; http://www.aricept.com/docs/pdf/aricept_PI.pdf; accessed 10 October 2015). Because the mean body weight in the present study was approximately 52 kg, higher plasma exposure might potentially cause decreased appetite with higher incidence compared with Study 326. Consequently, more patients than in Study 326 might show a weight decrease of ≥7% from baseline. Therefore, as reported by Han et al. [14], caution is required when prescribing 23 mg/day donepezil to patients with a lower body weight (e.g., Asians) because they may be more susceptible to AEs. Furthermore, those apparent or latent “not energetic” conditions might have a negative impression on caregivers and those who rate the CIBIC-plus. Because evidence suggests that stepwise titration may reduce the frequency of common AEs, such as nausea, vomiting, and diarrhea (Aricept Package,US; http://www.aricept.com/docs/pdf/aricept_PI.pdf; accessed 10 October 2015), the donepezil dose was escalated over 4 weeks in our study. A dose of 15 mg donepezil was estimated to be the median dose between 10 mg/day and 23 mg/day, and it was estimated that the steady-state of donepezil levels would be reached after 2 weeks. The donepezil dose in the SR 23 mg group in our study was initially increased from 10 mg/day to 15 mg/day after randomization, and then increased to 23 mg/day after 4 weeks of treatment. By contrast, the donepezil dose in Study 326 was increased from 10 mg/day to 23 mg/day immediately afterrandomization.

Table 4 shows the incidences of AEs (nausea, vomiting, and diarrhea) in Studies 343 and 326 [15]. The incidence of any AEs was comparable in both studies. The incidences of nausea, vomiting, and diarrhea in the 10 mg/day group were also comparable between both studies. However, the incidence of nausea in the SR 23 mg group in Study 343 was 50% lower than that in Study 326. The comparison between different studies should be interpreted carefully. Also, in a previous study of patients with SAD, the frequency of nausea was lower than that of vomiting [3]. Therefore, we cannot exclude the possibility that the frequency of nausea in Study 343 was lower than that in Study 326 because the majority of patients in Study 343 were severely impaired and could not complain of subjective symptoms such as nausea. However, the findings suggest that a stepwise dose escalation strategy might improve the tolerability of the higher donepezildose.

The inhibition of AChE may induce the activation of parasympathetic tone. Donepezil is an AChE inhibitor that might influence QT interval due to bradycardia by accentuation of the parasympathetic nervous system (Aricept Package insert, Japan; http://www.eisai.jp/medical/products/di/EPI/ART_T-FG_EPI.pdf.; accessed 10 October 2015). Although the cardiac effects of donepezil are not fully understood, overdose of this drug may induce QT prolongation. The present study evaluated an effect of SR 23 mg/day donepezil on QT interval using standardized ECG equipment and at a central laboratory (100 Overlook Center, Princeton, NJ, USA) according to a standard procedure. Regarding the mean variation of QTcF in the SR 23 mg group, a prolongation of ≥5 ms from baseline was observed at each point during the 24 weeks. This value was defined as the “mean value of action on QTc with which authorities are concerned” according to ICH guidelines. The proportions of patients with a maximum QTcF of >450 ms and with a maximum ΔQTcF of >30 ms were higher in the SR 23 mg group than in the IR 10 mg group. There was no notable difference in the proportions of patients with a maximum QTcF of >500 ms and patients with a maximum ΔQTcF of >60 ms between the two groups. In addition, there was no difference in the incidences of AEs (Preferred Term: ECG QT prolonged) related to QT prolongation between the two groups. In the SR 23 mg group, six patients had a maximum ΔQTcF of >60 ms or a maximum QTcF of >500 ms. Of these, five patients had AEs related to QT prolongation and the other patient had a prolongation of > Δ60 ms after study withdrawal. Their severities were judged as mild or moderate, but not serious. In general, drug-induced QT prolongation is more often observed in women than men. Of note, in the present study, all six patients with AEs related to QT prolongation were women.

QTcF prolongation of >500 ms, which is defined as “QTc prolongation specially to be concerned” in ICH guidelines, was observed in only one patient in the SR 23 mg group at Week 24. This QTcF prolongation was evaluated as not related to the study drug by the investigator, and its severity was judged as mild. This patient had a baseline QTcF of 443 ms and was considered to have had preexisting QT prolongation. Serum concentrations of potassium in this patient were within standard ranges from baseline to Week 12 and values at Weeks 18 and 24 when QTcF >500 ms were all below the lower limit of standard values (3.5 mEq/L) (3.3 mEq/L and 3.0 mEq/L at Weeks 18 and 24, respectively). The results of the present study suggest a tendency for QTcF prolongation after increasing the donepezil dose from 10 mg/day to 23 mg/day. QTcF prolongations in the SR 23 mg group were not clinically significant. Therefore, it was suggested that the effect of SR 23 mg/day donepezil on QTcF was within the range that could be expected to maintain the good safety profile of donepezil, and therefore did not indicate a new risk.

CONCLUSIONS

A previous study of Japanese patients with SAD revealed dose-dependent responses of donepezil at doses of 0 mg/day (i.e., placebo), 5 mg/day, and 10 mg/day [2], leading to the approval of 10 mg/day donepezil as the highest approved dose in Japan. In the present 24-week study of Japanese patients with SAD treated with 10 mg/day donepezil for ≥3 months before enrollment, cognitive functions and global clinical changes did not improve in the 23 mg/day group relative to the 10 mg/day group. Moreover, SR 23 mg/day donepezil was not statistically superior to IR 10 mg/day donepezil in terms of the primary endpoints of this study (SIB and CIBIC-plus). Although no new safety concerns occurred, the AE incidence was greater in the SR 23 mg group than in the IR 10 mg group. Based on the current study results, the suggested optimum dose of donepezil for Japanese patients with SAD is 10 mg/day. However, future studies should investigate donepezil doses between 10–23 mg/day to determine the optimum dosage.

Footnotes

ACKNOWLEDGMENTS

The authors thank the clinical investigators and their institution staff, and the subjects and caregivers who participated in this clinical study (E2020-J081-343). The authors thank Oneeb Majid, PhD, of Eisai Ltd., Hatfield, UK, for performing population pharmacokinetic analysis. The authors also thank J. Ludovic Croxford, PhD, and Nicholas D. Smith, PhD, of Edanz Group Limited for providing medical writing support.

Authors’ disclosures available online ().

Appendix

References

Howard

, McShane

, Lindesay

, Ritchie

, Baldwin

, Barber

, Burns

, Dening

, Findlay

, Holmes

, Hughes

, Jacoby

, Jones

, McKeith

, Macharouthu

, O’Brien

, Passmore

, Sheehan

, Juszczak

, Katona

, Hills

, Knapp

, Ballard

, Brown

, Banerjee

, Onions

, Griffin

, Adams

, Gray

, Johnson

, Bentham

, Phillips

(2012) Donepezil and memantine for moderate-to-severe Alzheimer’s disease. N Engl J Med 366, 893–903.

Atri

, Hendrix

, Pejović

, Hofbauer

, Edwards

, Molinuevo

, Graham

(2015) Cumulative, additive benefits of memantine-donepezil combination over component monotherapies in moderate to severe Alzheimer’s dementia: A pooled area under the curve analysis. Alzheimers Res Ther 7, 28.

Homma

, Imai

, Tago

, Asada

, Shigeta

, Iwamoto

, Takita

, Arimoto

, Koma

, Ohbayashi

(2008) Donepezil treatment of patients with severe Alzheimer’s disease in a Japanese population: Results from a 24-week, double-blind, placebo-controlled, randomized trial. Dement Geriatr Cogn Disord 25, 399–407.

Mega

, Cummings

, Fiorello

, Gornbein

(1996) Spectrum of behavioral changes in Alzheimer’s disease. Neurology 46, 130–135.

Finkel

, Costa e Silva

, Cohen

, Miller

, Sartorius

(1996) Behavioral and psychological signs and symptoms of dementia: A consensus statement on current knowledge and implication for research and treatment. Int Psychogeriatr 8(Suppl 3), 497–500.

Kaasinen

, Någren

, Järvenpää

, Roivainen

, Yu

, Oikonen

, Kurki

, Rinne

(2002) Regional effects of donepezil and rivastigmine on cortical acetylcholinesterase activity in Alzheimer’s disease. J Clin Psychopharmacol 22, 615–620.

Farlow

, Salloway

, Tariot

, Yardley

, Moline

, Wang

, Brand-Schieber

, Zou

, Hsu

, Satlin

(2010) Effectiveness and tolerability of high-dose (23 mg/d) versus standard-dose (10 mg/d) donepezil in moderate to severe Alzheimer’s disease: A 24-week, randomized, double-blind study. Clin Ther 32, 1234–1251.

Schmitt

, Ashford

, Ernesto

, Saxton

, Schneider

, Clark

, Ferris

, Mackell

, Schafer

, Thal

(1997) The severe impairment battery: Concurrent validity and the assessment of longitudinal change in Alzheimer’s disease. The Alzheimer’s Disease Cooperative Study. Alzheimer Dis Assoc Disord 11, S51–S56.

Joffres

, Graham

, Rockwood

(2000) Qualitative analysis of the clinician interview-based impression of change (Plus): Methodological issues and implications for clinical research. Int Psychogeriatr 12, 403–413.

10.

Schneider

, Olin

(1996) Clinical global impressions in Alzheimer’s clinical trials. Int Psychogeriatr 8, 277–288.

11.

Kuhl

, Minoshima

, Frey

, Foster

, Kilbourn

, Koeppe

(2000) Limited donepezil inhibition of acetylcholinesterase measured with positron emission tomography in living Alzheimer cerebral cortex. Ann Neurol 48, 391–395.

12.

US Food and Drug Administration (2010) Clinical Pharmacology Biopharmaceutics Review(s). Drug Approval Package. http://www.accessdata.fda.gov/drugsatfda_docs/nda/2010/022568Orig1s000ClinPharmR.pdf. Posted 1 December 2010, Accessed 3 January 2016.

13.

Sabbagh

, Cummings

, Christensen

, Doody

, Farlow

, Liu

, Mackell

, Fain

(2013) Evaluating the cognitive effects of donepezil 23 mg/d in moderate and severe Alzheimer’s disease: Analysis of effects of baseline features on treatment response. BMC Geriatr 13, 56.

14.

Han

, Lee

, Kubota

, Dash

(2014) Efficacy and tolerability of donepezil 23 mg/d versus donepezil 10 mg/d in patients with moderate to severe AD in an Asian population: Study 326 subanalysis. Alzheimers Dement 10, 450–451.

15.