Memantine benefits functional abilities in moderate to severe Alzheimer's disease

Abstract

Objective

Functional abilities are severely impacted in Alzheimer's disease (AD). Loss of the ability to perform complex (instrumental) and basic activities of daily living (ADL), leads to decreased independence and increased caregiver burden. This post-hoc analysis investigated the effect of memantine (20 mg/day) on ADLs, as measured by Alzheimer's Disease Cooperative Study-Activities of Daily Living 19-item (ADCSADL₁₉) and 23-item (ADCS-ADL₂₃) scales, in patients with moderate-to-severe AD.

Design

Data were pooled from six multicenter, randomized, placebo-controlled, double-blind, 6-month studies of memantine 20 mg/day.

Participants

Male and female patients aged ≥50 years at baseline with a Mini Mental State Examination (MMSE) score <20.

Measurements

ADCS-ADL₁₉ and ADCS-ADL₂₃ scales were pooled, and 14 shared items, with a score range of 0–45, were identified and included in the analysis (ADL₁₄). Basic ADLs (BADLs) were defined as: eating, walking, toileting, bathing, and grooming. Instrumental ADLs (IADLs) were defined as: using a telephone, watching television, conversing, clearing a table, finding belongings, obtaining a beverage, disposing of household rubbish, travelling outside the house, and being left alone. Changes from baseline on single-item, BADL (range: 0–15), IADL (range: 0–30), and total ADL₁₄ scores were analysed for observed cases using ANCOVA, with study, center and treatment as categorical explanatory variables and score at baseline as a covariate.

Results

959 patients were treated with memantine and 867 patients received placebo. Memantine-treated patients had less decline from baseline on the ADL₁₄ total score, compared with placebo (p<0.001) at study end. Memantine also showed lower reductions in BADLs (p<0.05) and IADLs (p<0.001), for observed cases, compared with placebo. Memantine-treated patients showed less worsening than placebo recipients for the ADL items: toileting (p<0.01), grooming (p<0.01), finding belongings (p<0.01), and travelling outside the house (p<0.05).

Conclusion

Compared with placebo, memantine shows benefits for both basic and instrumental ADLs in patients with moderate-to-severe AD, suggesting that memantine treatment may lead to a more interactive and dignified life for patients with moderate-to-severe AD.

Introduction

Alzheimer's disease (AD) is a progressive neurodegenerative condition characterized by memory impairment, deterioration of language, other cognitive deficits, and behavioural problems such as confusion, irritability, aggression, mood swings and general withdrawal (1). In addition, functional impairments include an inability to perform instrumental daily activities and basic daily activities (2). It is the single most common cause of dementia, and in Europe, accounts for approximately 55-80% of all cases of dementia (3).

Progression of AD is defined by increasing functional and behavioural impairments, which, in the moderate to severe stages of the disease, can result in total dependence on a caregiver and a high risk of institutionalisation (4, 5). Thus, a patient's functional performance gradually progresses from an inability to perform instrumental daily activities, such as telephoning and travelling alone, to an eventual decline in the basic daily activities of eating, walking or grooming (2). Deterioration in functional performance can be measured over the space of 6-12 months (2). This loss of functional ability is also one of the biggest problems for caregivers and contributes towards the major burden for carers (6).

Treatment interventions that enable a patient with AD to retain some ability to perform activities of daily living and to preserve their independence for longer, may prevent increases in patient distress and caregiver burden that would ultimately lead to patient institutionalisation. Memantine, a specific, uncompetitive N-methyl-d-aspartate (NMDA) receptor antagonist with moderate affinity and rapid voltage-dependent kinetics, acts by targeting the neurotransmitter (glutamate) system central to the pathology of AD (7, 8). Memantine has been approved in Europe and the USA for the treatment of patients with moderate-to-severe AD, and is the only treatment in clinical use for AD that specifically targets the glutamatergic system.

In clinical studies of outpatients with AD, memantine had significant clinical effects on global status, cognitive (including language, memory, orientation, praxis, and visuospatial ability), functional, and behavioural (including symptoms of delusions, agitation/aggression and irritability/lability) outcomes, and was well tolerated with an adverse event profile similar to placebo (9, 10, 11, 12). Furthermore, in nursing home patients with severe dementia (AD or vascular dementia (VaD)), memantine improved activities of daily living, care dependence and behavioural symptoms compared with placebo (13). For these reasons, memantine has the potential to improve the functional status of patients to a greater extent than that seen with donepezil, for which there is conflicting evidence in relation to its benefits on activities of daily living in severe AD (14, 15, 16).

Six randomized, placebo controlled trials have assessed the efficacy and safety of memantine 20 mg/day (17, 18, 19, 20, 21, 22) and the data on cognitive outcomes, behavioural symptoms and safety and tolerability from these trials have been assessed in detail in meta-analyses (9, 10, 12). The information on functional performance from these studies has not been assessed in the same way, and therefore, in this post hoc analysis, data from these six studies were pooled to evaluate the effects of treatment on functional performance in patients with moderate-to-severe AD.

Methods

Study design and patient population

Data were pooled from six multicenter, randomized, placebo-controlled, double-blind, 6-month studies of memantine 20 mg/day (Table 1). All English-language published double-blind, randomized, placebo-controlled, phase III studies of memantine with a minimum duration of 24 weeks were included in the analysis. All patients, or their legally authorized representatives, provided written informed consent prior to participation.

Table 1.

Summary of the six memantine (20 mg/day) randomized, double-blind, placebo-controlled studies included in this analysis

Study	MMSE inclusion range (mean)	Duration and other details	Patients	Efficacy measures
Peskind et al (2006)	10-22 (17.3) – mild to moderate	24 weeks Twice-daily memantine	403 patients P: 202 M: 201	ADAS-cog, ADCS-ADL23, CIBIC-Plus, NPI
Porsteinsson et al (2008)	10-22 (16.9) – mild to moderate	24 weeks, patients already receiving a cholinesterase inhibitor Once-daily memantine	433 patients P: 216 M: 217	ADCS-ADL23, CIBIC-Plus, MMSE, NPI
Bakchine and Loft (2008)	11-23 (18.7) – mild to moderate	24 weeks Twice-daily memantine	470 patients P: 152 M: 318	ADAS-cog, ADCS-ADL23, CIBIC-Plus, NPI
Reisberg et al (2003)	3-14 (7.9) – moderate to severe	28 weeks Twice-daily memantine	252 patients P: 126 M:126	SIB, ADCS-ADL19, CIBIC-Plus, FAST, GDS, MMSE, NPI
Van Dyck et al (2007)	5-14 (10.1) – moderate to severe	24 weeks Twice-daily memantine	350 patients P: 172 M: 178	SIB, ADCS-ADL19, BGP, CIBIC-Plus, FAST, NPI
Tariot et al (2004)	5-14 (10.0) – moderate to severe	24 weeks, patients already receiving donepezil Twice-daily memantine	403 patients P: 201 M: 202	ADCS-ADL19, BGP, CIBIC-Plus, NPI, SIB

P = placebo; M = memantine; ADAS-cog = Alzheimer's Disease Assessment Scale Cognitive Subscale; SIB = Severe Impairment Battery; ADCS-ADL₁₉ = 19-item AD Cooperative Study-Activities of Daily Living Inventory; ADCS-ADL₂₃ = 23-item AD Cooperative Study-Activities of Daily Living Inventory; BGP = Behavioural Rating Scale for Geriatric Patients; CIBIC-plus = Clinician's Interview-Based Impression of Change Plus Caregiver Input; FAST = Functional Assessment Staging; GDS = Global Deterioration Scale; NPI = Neuropsychiatric Inventory. Cholinesterase inhibitors were donepezil, rivastigmine, or galantamine.

Study design and patient populations have been described in detail elsewhere (17, 18, 19, 20, 21, 22). All patients were diagnosed with probable AD according to the NINCDS-ADRDA criteria (23, 24). Briefly, study participants were male or female outpatients aged =50 years at baseline. In all studies, patients received an initial memantine dose of 5 mg/day (or placebo), which was titrated in 5 mg/day increments at weekly intervals to 20 mg/day. Concomitant psychotropic medications were permitted in five studies (17, 18, 20-22), but not in one study (19).

Efficacy outcome measures

Functional abilities were assessed in all six studies (17, 18, 19, 20, 21, 22) using the AD Cooperative Study-Activities of Daily Living 19-item (ADCS-ADL₁₉) or 23-item (ADCS-ADL₂₃) structured questionnaires at baseline, and after 4, 12 and 24 weeks. The ADCS-ADL₁₉ is used for patients with moderate to severe AD and the ADCS-ADL₂₃ for patients with mild to moderate AD (11). Each item on these scales evaluates a patient's ability to perform an activity of daily living. Caregivers assess a patient's performance during the previous four weeks; increasing scores indicate an improvement in functional ability, with a total score of 54 (for ADCS-ADL₁₉) or 78 (for ADCS-ADL₂₃) indicating optimal performance. In the present pooled analysis, mean change in ADL score from baseline was assessed for 14 overlapping items on the ADCS-ADL₁₉ and ADCS-ADL₂₃ scales (for the purposes of this paper, referred to as ADL₁₄), because three studies recruited patients with mild-to-moderate AD and therefore used the ADCS-ADL₂₃ (17, 18, 22) and three studies recruited patients with moderate-to-severe AD and therefore used the ADCS-ADL₁₉ (19, 20, 21). The score range for the ADL₁₄ was 0-45. The basic ADLs (BADLs) comprised the items eating (score range 0–3), walking (0–3), toileting (0–3), bathing (0–3), and grooming (0–3), and the instrumental ADLs (IADLs) were the remaining non-basic items, which were using a telephone (0–5), watching television (0–3), conversing (0–3), clearing a table (0–3), finding belongings (0–3), obtaining a beverage (0–3), disposing of household rubbish (0–3), travelling outside the house (0–4), and being left alone (0–3). For each item, scores ranged from 0 = ‘least able to perform task' to 3 or 4 = ‘most able to perform task'.

Statistical analyses

Efficacy analyses were conducted on the patients from the full analysis set (FAS; all patients who received at least one dose of study medication and had at least one post-baseline efficacy assessment) with Mini Mental State Examination (MMSE) score <20 (moderate-to-severe AD), using an observed-case (OC) approach and a last-observation-carried forward (LOCF) approach to support the OC analyses. The efficacy endpoints were the change from baseline in ADL₁₄ total score, BADL score and IADL score, and change in ADL₁₄ single-item scores. Differences between memantine and placebo were analyzed within the OC and LOCF populations using analysis of covariance (ANCOVA), with study, center, and treatment as categorical explanatory variables and score at baseline as a covariate. All efficacy analyses were tested at the two-sided 95% level of significance.

Results

Baseline characteristics

The FAS consisted of the 959 memantine-treated patients and 867 placebo recipients who had baseline MMSE scores of less than 20. Baseline demographic and clinical characteristics, including scores for the ADCS-ADL₁₉, ADCS-ADL₂₃, and ADL₁₄, were similar in the two treatment groups (Table 2).

Table 2.

Patient demographics and clinical characteristics at baseline (FAS; MMSE <20)

	Placebo (n=867)	Memantine (n=959)
Women, n (%)	550 (63.4)	644 (67.2)
Mean (±SD) age, years	76.2 ±8.3	76.2 ±8.1
Caucasian ethnicity, n (%)	788 (90.9)	865 (90.2)
MMSE score, mean ± SD	12.2 ± 4.1	12.3 ± 4.2
ADL19 total score, mean ± SD	33.1± 10.7(n=499)	32.5±10.7(n=506)
ADL23 total score, mean ±SD	52.7 ± 13.2 (n=368)	51.3 ± 14.7 (n=453)
ADL14 total score, mean ±SD	29.3 ± 8.5 (n=850)	29.0 ± 8.7 (n=938)
Basic ADLs	12.4 ± 2.9 (n=850)	12.4 ± 3.0 (n=938)
Eating	2.7±0.5(n=850)	2.7±0.5(n=938)
Walking	2.9±0.5(n=849)	2.8±0.5(n=938)
Toileting	2.6±0.8(n=850)	2.6±0.8(n=938)
Bathing	2.0±1.0(n=850)	2.0±1.0(n=938)
Grooming	2.3±1.0(n=849)	2.2±1.0(n=937)
Instrumental ADLs	16.8 ± 6.2 (n=850)	16.6 ± 6.3 (n=938)
Using a telephone	2.1±1.3(n=850)	2.1±1.3(n=938)
Watching television	1.0±1.0(n=835)	1.0±1.0(n=925)
Conversing	2.1±1.2(n=850)	2.1±1.1(n=937)
Clearing a table	2.3±1.1(n=850)	2.2±1.2(n=938)
Finding belongings	1.9±1.1(n=850)	1.9±1.1(n=938)
Obtaining a beverage	1.6±1.2(n=850)	1.6±1.2(n=937)
Disposing of litter	2.4±1.0(n=850)	2.4±1.1(n=938)
Travelling outside the house	2.4±0.9(n=850)	2.3±0.9(n=938)
Being left alone	1.5±1.2(n=540)	1.6±1.2(n=619)

ADL₁₉ and ADL₂₃ = Alzheimer's Disease Cooperative Study-Activities of Daily Living 19 and 23 item scales; ADL₁₄ = overlapping items on the ADL₁₉ and ADL₂₃ scales; MMSE = Mini Mental State Examination.

Functional efficacy

Memantine-treated patients had smaller reductions (i.e., less worsening) in scores on the ADL₁₄ than placebo recipients. Differences in functional decline between memantine-treated and placebo-treated patients increased over time (Figure 1).

Figure 1.

Mean change from baseline in ADL₁₄ total score, at Weeks 4, 12 and 24/28 (observed cases) and at Week 24/28 (last observation carried forward) in the pooled population of patients with moderate to severe AD. P-values from ANCOVA

BADLs and IADLs decreased by less in the memantine-treated patients than in placebo-treated patients (Figure 2). The results from the LOCF analysis were consistent with those of the OC analysis: at week 4 there were no significant differences in the BADL or IADL scores in the two LOCF patient groups, but by week 12 IADL scores were significantly less reduced in memantine-treated patients than placebo-treated patients, and by week 24 both BADL and IADL were significantly less reduced with memantine. Significantly less worsening of individual ADL scores (two BADL single-item scores [toileting and grooming], and two IADL single-item scores [finding belongings and travelling outside the house]) were also recorded at week 24 in memantine recipients, compared with placebo (Figure 3). The LOCF analysis showed similar significant differences, except the difference between groups for the travelling outside the house item was not significant (p=0.14). Conversing worsened more in the placebo-group than in the memantine-treated group, but the difference did not reach statistical significance (p=0.06 in OC analysis and p=0.08 in LOCF analysis).

Figure 2.

Mean change from baseline in ADL₁₄ BADL and IADL score at Week 24/28 (observed cases) in the pooled population of patients with moderate to severe AD. P-values from ANCOVA

Figure 3.

Mean change from baseline in ADL single-item scores at Week 24/28 (observed cases)

Discussion

In this pooled post hoc analysis of six 6-month studies of patients with moderate to severe AD (MMSE<20), memantine (20 mg/day) treatment provided statistically significant benefits in overall functional ability and activities of daily living compared with placebo. These results confirm the initial, less comprehensive, report of better functional ability with memantine than placebo (11), and suggest that memantine treatment may help patients maintain a more interactive life. The ADCS-ADL₁₉ and ADCS-ADL₂₃ are reliable, sensitive and validated instruments for assessing functional change in patients with moderate to severe AD (2).

Although the patients recruited in the clinical trials analysed in this study were not selected for having functional impairments, mean ADL₁₄ total scores in both memantine and placebo treatment groups at baseline were indicative of patients with impaired functional ability. At study end, mean total score significantly favoured memantine and a single-item analysis revealed the significant benefits of memantine, compared to placebo, in improving patients' abilities to perform basic daily activities, such as toileting and grooming, which are often impaired in the moderate to severe stages of the disease (2). Analyses according to OC and LOCF approaches did not markedly influence the observed results in this analysis, indicating the consistency of the functional benefits shown with memantine. Furthermore, significant benefits of memantine therapy were seen as early as week 12 (according to the ADL₁₄ total scores and IADL scores).

The six trials included in this analysis all recruited outpatients, but the findings were also congruent with results from the subset analysis of nursing home patients with AD in the 9M-BEST study (13).

An assessment of clinical effectiveness showed that combination therapy with memantine plus cholinesterase inhibitors slowed both cognitive and functional decline (measured using the Weintraub ADL Scale) compared with cholinesterase inhibitor monotherapy or no treatment (25, 26). In the current analysis, differences in functional decline between memantine-treated and placebo-treated patients increased over time, and Atri et al showed similar increases with time that were sustained for at least four years (predicted frequency of dependence rising from 37% to 74% in untreated patients and from 32% to 57% in patients treated with memantine plus cholinesterase inhibitors) (25).

AD is a complex disease, with a progressive course that manifests as disturbances across functional, behavioural and cognitive domains (1). As the only approved treatment for AD that targets the glutamatergic system, memantine has been shown to confer statistically significant advantages compared with placebo on cognitive, behavioural, and global endpoints throughout the course of AD (10-12, 26). In addition, the current analysis provides evidence of benefits for memantine on functional outcomes based on six 6-month studies in moderate to severe AD. Furthermore, in efficacy and safety studies, memantine treatment was consistently well tolerated (17, 18, 19, 20, 21, 22). Indeed, the overall incidence of treatment-related adverse events with memantine did not differ from placebo, in a published pooled analysis of six randomized trials of memantine 20 mg/day in patients with AD (11).

It is well established that caregiver burden increases, as does the risk of institutionalisation, in the more advanced stages of AD (4, 5) – and functional decline plays a major part in the burden and risk. The results from this post hoc analysis demonstrate that memantine supports patient independence by reducing the decline in basic and instrumental skills and their ability to perform everyday tasks. This could potentially ease the burden of caregiving and improve patient and carer quality of life, and one study has shown that patients treated with memantine required 52 hours less caregiver time per month (1.5 h/day), than those receiving placebo (27). In addition, patients with AD receiving memantine were three times more likely to remain autonomous after 28 weeks of daily treatment than patients receiving placebo (28), and recently it was shown that nursing home admission for patients with AD was significantly delayed during memantine treatment, in combination with acetylcholinesterase inhibitors, compared with patients receiving acetylcholinesterase inhibitors alone or receiving no cognitive enhancers (29).

The investigation of single-items may lead to the identification of specific areas of treatment effects; however, it can also increase the risk of identifying random performance fluctuations. The results of the single-item/sub-scale analyses identified specific items for which memantine demonstrated a significant beneficial effect, and these results were consistent at the analysed time-points. The consistent results indicate that the single-item/sub-scale scores are not random fluctuation but significant effects of memantine on specific sub-domains. There are, however, limitations to this study as it is a post-hoc study with no correction for multiple testing.

These results suggest that memantine treatment may lead to a more interactive and dignified life for patients with moderate-to-severe AD, than placebo. The functional improvements observed in this analysis add to the growing evidence supporting the clinically meaningful benefits of memantine in patients with moderate-to-severe AD.

Acknowledgements: The authors would like to thank all patients and caregivers participating in the studies, and thank all investigators for their contributions to the studies. H. Lundbeck A/S, Merz Pharmaceuticals GmbH and Forest Research Institute Inc. funded the studies used for this analysis. H. Lundbeck A/S conducted the pooled analysis and editorial assistance for this manuscript was provided by ESP Ltd (Sandhurst, UK) with funding from H. Lundbeck A/S.