Safety and efficacy of galantamine (Reminyl) in severe Alzheimer s disease (the SERAD study): a randomised, placebo-controlled, double-blind trial

Transcription

1 Safety and efficacy of galantamine (Reminyl) in severe Alzheimer s disease (the SERAD study): a randomised, placebo-controlled, double-blind trial Alistair Burns, Roberto Bernabei, Roger Bullock, Alfonso J Cruz Jentoft, Lutz Frölich, Christoph Hock, Minna Raivio, Eric Triau, Maurits Vandewoude, Anders Wimo, Elizabeth Came, Bart Van Baelen, Gerry L Hammond, Joop C van Oene, Susanne Schwalen Summary Background The efficacy of galantamine has been shown in patients with mild, moderate, and advanced moderate Alzheimer s disease (AD). Here we report its efficacy in patients with severe AD. Methods Between December, 2003, and March, 2007, patients aged 84 (SD 6) years with severe AD (mini-mental state examination [MMSE] score 5 12 points), in a nursing home setting were randomly assigned to receive galantamine (n=207), titrated initially to 24 mg/day, or placebo (n=200). Co-primary efficacy measures for cognitive function and ability to undertake normal daily activities were the severe impairment battery (SIB) and the seven-item minimum data set-activities of daily living (MDS-ADL), respectively. Adverse events, vital signs, laboratory parameters, and electrocardiograms were monitored. This trial is registered with ClinicalTrials.gov, number NCT Findings 168 of 207 (81%) patients in the galantamine group and 161 of 200 (81%) in the placebo group completed the study. Mean SIB scores increased (improved) by 1 9 (95% CI 0 1 to 3 9) points with galantamine and decreased (worsened) by 3 0 ( 5 6 to 0 5) points with placebo (between-group least squares mean difference 4 36, 1 3 to 7 5; p=0 006). Mean MDS-ADL self-performance score worsened by 1 2 (0 6 to 1 8) points and 1 6 (0 8 to 2 3) points, respectively (between-group least squares mean difference 0 41, 1 3 to 0 5; p=0 383). Nominally significant between-group differences in favour of galantamine occurred for the SIB domains of memory (p=0 006), praxis (p=0 010), and visuospatial ability (p=0.002), and for the MDS-ADL subitem locomotion on unit (p=0 021). 183 of 207 patients (88%) who received galantamine and 177 of 200 (89%) who received placebo had adverse events, which were mostly mild to moderate. Eight patients (4%) in the galantamine group and 21 patients (11%) in the placebo group died. ECG abnormalities were similar between the two groups. Interpretation can be started and used safely in elderly patients with severe AD. improved cognitive function but failed to significantly improve the co-primary parameter of overall activities of daily living. Funding Janssen-Cilag EMEA. Introduction Alzheimer s disease (AD) is an age-related neurodegenerative disorder that is increasing in prevalence as the population lives longer. 1 For diagnostic, treatment, and research purposes, various stages of Alzheimer s disease (AD) can be defined to indicate the level of impairment during progression from mild to moderate and severe to terminal dementia. As the disease progresses, cognitive decline extends to include impairments in the ability to undertake basic activities of daily living, such as walking without assistance, bathing, and dressing. 2 Thus, as the severity of AD increases, so too does the burden on caregivers, which is an important predictor of the requirement for long-term inpatient care. 3 Furthermore, the provision of adequate care for patients with severe AD carries enormous costs. 4 is a reversible inhibitor of cholinesterase and an allosteric-potentiating ligand of the nicotinic acetylcholine receptors. 5 Clinical trials of inhibitors of cholinesterase have focused on patients with mild and moderate AD; therefore, efficacy data in patients with severe AD are scarce. Patients with milder disease will, however, progress to severe AD, and 20% of patients with AD are estimated to have severe dementia at any given time. 6 The benefits of galantamine on cognitive function, activities of daily living, and behaviour in patients with mild to moderate AD have been shown in randomised controlled trials and confirmed in meta-analyses. 7,8 The efficacy of galantamine was also assessed in a subpopulation of patients with advanced moderate AD (baseline mini-mental state examination [MMSE] 9 score points), with significant benefits on cognitive function compared with placebo, maintenance of functional ability, and delay of behavioural symptoms. 10 Similar benefits were seen in a pooled post hoc analysis of data from 12 months follow-up in two studies of patients with advanced moderate AD (MMSE 14 points or Alzheimer s disease assessment scale-cognitive subscale [ADAS-cog] score >30 points). 11 Significant benefits have been reported on cognitive function, activities of daily living, and behaviour with donepezil, an inhibitor of cholinesterase with no nicotinic Lancet Neurol 2009; 8: Published Online December 1, 2008 DOI: /S (08) See Reflection and Reaction page 39 University of Manchester, Manchester, UK (A Burns MD); Università Cattolica del Sacro Cuore, Rome, Italy (R Bernabei MD); Kingshill Research Centre, Swindon, UK (R Bullock MD); Hospital Universitario Ramón y Cajal, Madrid, Spain (A J Cruz Jentoft MD); Central Institute for Mental Health, Mannheim, Germany (L Frölich MD); University of Zürich, Zürich, Switzerland (C Hock MD); Memory Research and Treatment Centers Finland-Medical Center Hemo, Lahti, Finland (M Raivio MD); RVT De Wingerd, Leuven, Belgium (E Triau MD); University of Antwerp, Antwerp, Belgium (M Vandewoude MD); Primary Care Nordanstig and ADRC, Karolinska Institutet, Stockholm, Sweden (A Wimo MD); Came Clinical Research, Devizes, UK (E Came PhD); SGS Life Sciences Services, Mechelen, Belgium (B Van Baelen MSc); Janssen- Cilag EMEA, Beerse, Belgium (G L Hammond BSc); Janssen- Cilag EMEA, Tilburg, Netherlands (J C van Oene PhD); University Witten/Herdecke, Witten, Germany (S Schwalen MD) Correspondence to: Alistair Burns, University of Manchester, University Place, Manchester M13 9PL, UK alistair.burns@manchester. ac.uk Vol 8 January

2 505 screened* 407 enrolled 207 assigned to galantamine 200 assigned to placebo 39 discontinued 30 had adverse events 5 for non-compliance 2 withdrew consent 2 for other reasons 98 failed screening 168 completed trial 161 completed trial 39 discontinued 31 had adverse events 2 for non-compliance 3 withdrew consent 3 for other reasons Figure 1: Trial profile *Excluding eight patients from one site that was closed prematurely. Four patients from a second site were also excluded after screening, but before randomisation. To be eligible for the intention-to-treat population, patients required data for both primary efficacy parameters (galantamine n=186; placebo n=178). Other defined as moved to a new permanent residential setting (n=3); patient unwilling to cooperate (n=1); and patient did not meet inclusion criteria (n=1). receptor-modulating properties compared with placebo in patients with moderate-to-severe AD. 12 Recently, a randomised controlled study of patients with severe AD (MMSE 1 10 points) in Swedish nursing homes found that donepezil significantly improved cognitive function and prevented decline in activities of daily living compared with placebo. 13 Similarly, in a study of patients with advanced AD (MMSE 1 12 points) donepezil was superior to placebo with regard to effects on cognitive function and global function but did not differ significantly from placebo on activities of daily living. 14 The SERAD trial was designed to assess the safety of galantamine and its efficacy on cognitive function and activities of daily living in patients with severe AD (MMSE 5 12 points) in nursing homes across Europe. Methods Participants Patients aged years who were diagnosed with dementia of the Alzheimer type (Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition 15 ) or probable Alzheimer s dementia (NINCDS-ADRDA classification 16 ) were recruited. Patients with possible Alzheimer s dementia with cerebrovascular disease (NINDS-AIREN criteria 17 ) were also eligible if their AD was thought to be the most probable cause of the progressive dementia. Inclusion criteria were severe dementia (MMSE 5 12 points), a history of cognitive decline during at least 6 months, and a brain CT or MRI scan within 3 years before enrolment. Patients were required to be ambulatory and have sufficient (assisted) vision and hearing to comply with testing. Previous cholinesterase inhibitor or memantine therapy must have been discontinued at least 3 months before randomisation. Patients had to be resident for at least 3 months in a residential home, nursing home, or geriatric residence that provides long-term care for elderly people and people with dementia and employs paid, trained professional caregivers to help with personal care. Patients functioning was assessed by an assigned caregiver or a nominated caregiver team member who represented the team s views. Consent was obtained from patients whenever possible and consent or assent was authenticated by the investigator or an authorised member of the investigational staff for all patients. An informed consent form was also signed by an appropriate relative, guardian, or legal representative, and the reporting caregiver also gave written consent for their own participation or role in the study. Exclusion criteria included dementia primarily caused by cerebrovascular disease, disturbances of consciousness, delirium, psychosis, severe aphasia, major sensorimotor impairment that precluded neuropsychological tests, neurodegenerative disorders (eg, Parkinson s disease or Huntington s chorea), or other conditions that could possibly result in impaired cognitive function (eg, cerebral trauma, hypoxia, vitamin deficiency, or metabolic disturbance). Patients with clinically significant cardiovascular disease who were expected to have limited participation in a 6-month trial were also excluded; so too were patients with a history of epilepsy, drug abuse, severe drug allergy, peptic ulcer, clinically significant psychiatric, hepatic, renal, pulmonary, metabolic, or endocrine disturbance, or urinary outflow obstruction. Procedures SERAD was a 6-month, double-blind, randomised, placebo-controlled trial done at 57 investigational sites in 10 European countries. Computer randomisation was used before the start of the study, with randomisation codes generated by the Janssen Research Foundation, Beerse, Belgium. Trial medications were randomised in blocks of four two galantamine and two placebo and numbered. Randomisation was stratified by baseline MMSE category and investigational site. Each investigational site received at least two blocks: one for patients with MMSE scores of 5 8 points and one for those with MMSE scores of 9 12 points, with the aim of recruiting a similar number of patients into the two severity groups. Eligible patients were assigned to the next available medication number within the block. The investigator was provided with a sealed randomisation code for each available medication number. Blinding was to be broken only if the patient s trial medication would affect specific emergency treatment. Breaking the blind resulted in mandatory withdrawal of the patient from the study. All randomisation codes, opened or sealed, were retrieved at the end of the study. At each visit, patients received enough blinded medication of their specific number to cover their intake 40 Vol 8 January 2009

3 (n=207) (n=200) Age (years) 83 7 (5 7) 83 5 (5 8) Female 167 (81%) 162 (81%) Weight (kg) 61 3 (11 5) 61 7 (12 7) Height (m) (8 7) (8 5) Body-mass index (kg/m2) 24 4 (4 2) 24 7 (4 8) Diagnosis Probable AD 170 (82%) 157 (79%) Possible AD with CVD 37 (18%) 43 (22%) Duration of illness (months) 55 6 (34 7) 52 0 (24 0) SIB score 66 8 (20 9) 67 8 (21 5) MDS-ADL self-performance score 11 9 (7 6) 12 6 (7 5) MMSE score 8 8 (2 4) 9 1 (2 4) MMSE strata (as randomised) 5 8 points 86 (42%) 84 (42%) 9 12 points 121 (59%) 116 (58%) Active disease at baseline Any active concomitant disease 205 (99%) 196 (98%) Cardiovascular 139 (67%) 140 (70%) Psychiatric 137 (66%) 111 (56%) Musculoskeletal 122 (59%) 110 (55%) Gastrointestinal 119 (58%) 106 (53%) Genitourinary 92 (44%) 75 (38%) Neurological 90 (44%) 70 (35%) Endocrine 52 (25%) 58 (29%) Dermatological 46 (22%) 48 (24%) Eyes 44 (21%) 43 (21%) Haematological 38 (18%) 35 (18%) Respiratory 32 (16%) 21 (11%) Ears, nose, and throat 24 (12%) 22 (11%) Allergic or immunological 17 (8%) 13 (7%) Data are mean (SD) or number (%). AD=Alzheimer s disease. CVD=cerebrovascular disease. MDS ADL=minimum data set activities of daily living. SIB=severe impairment battery. MMSE=mini-mental state examination. Table 1: Demographic and baseline characteristics until the next scheduled visit. Appropriate tablets of different strengths (4, 8, or 12 mg galantamine) or placebo tablets of identical appearance were provided to meet the titration schedule: 4 weeks of 4 mg twice per day, 4 weeks of 8 mg twice per day, and 4 weeks of 12 mg twice per day. After titration, the maintenance dose was kept at 12 mg twice per day but could be reduced to 8 mg at the same frequency for reasons of tolerability. Efficacy evaluations were done at baseline and at weeks 8, 12, and 26, or at premature discontinuation. The primary efficacy evaluations were the severe impairment battery (SIB) 18 and the composite seven-item minimum data set-activities of daily living (MDS-ADL) selfperformance scale (ie, the caregiver rates the patient s performance). 19,20 SIB and MDS-ADL evaluations were each rated separately by different observers on separate case-report forms. Mean change from baseline in SIB score Between-group least square mean difference 4 36 (95% CI ); p=0 006 Improved Worsened Secondary efficacy variables included subitems of the SIB and the full 11-item MDS-ADL. Adverse events and vital signs were assessed at each study visit. In addition, electrocardiogram (ECG) and laboratory safety tests were done at screening, and at weeks 8 and 26 (end of treatment) or at premature discontinuation. The trial was done in accordance with current International Conference on Harmonisation guidelines on good clinical practice (GCP) and applicable regulatory requirements and was approved by the ethics committees of the participating countries. During the course of the study, two of the 57 investigational sites, with eight randomly assigned patients, were disqualified because of non-compliance with GCP. Data from these patients were excluded from the analyses before unblinding. Statistical analysis For 90% power to detect a treatment difference in the SIB at the 5% significance level, 144 patients were needed per group; this figure was calculated on the basis of a trial of donepezil in which the SD of the change in SIB from baseline at week 26 was about 13 points, 21 and an expected treatment improvement of 5 points in SIB score over placebo. With the same significance criteria, and for an anticipated SD of 3 3 points in the seven-item MDS-ADL parameters and a difference between active treatment and placebo of 1 25 points, 148 patients per group were needed. A change of greater than one point is deemed clinically relevant by the authors of the MDS. To recruit about 290 patients who could be evaluated at week 26, and with an estimated dropout rate of 33% (a conservative figure that is based on our experience in past studies), Primary efficacy parameter * Baseline Time point in double-blind phase (week) Number at risk Figure 2: Cognitive function Change from baseline in severe impairment battery total score (intention-to-treat population). Bars indicate 95% CI for the comparison with baseline. *p= p= p=0 006 versus placebo. Vol 8 January

4 Mean change from baseline in SIB subscale score Mean change from baseline in MDS-ADL self-performance score (n=162) (n=149) * Social interaction Between-group least square mean difference 0 41 (95% CI 1 3 to 0 5); p=0 383 Worsened Memory Orientation Language Attention Praxis 430 patients needed to be randomised. Because the dropout rate during the trial was lower than anticipated, recruitment was stopped in the end stages of the ongoing trial after randomisation of 415 patients when it became clear that the estimate of 290 patients completing 6 months of treatment was achievable: continuation of recruitment would have subjected some patients to unnecessary procedures. The statistical analysis plan was finalised before the database was closed and unblinded. Analyses were done with SAS software by employees of SGS Life Science Services who were unaware of the blinding code. All patients who were randomised or treated were included in the safety analyses. Efficacy analyses were based on the Visuospatial ability Primary efficacy parameter 2 5 Baseline Time point in double-blind phase (week) Number at risk Figure 3: Activities of daily living Change from baseline in seven-item MDS ADL self-performance score (intention-to-treat population). Bars indicate 95% CI for the comparison with baseline. *p= p= p=0 383 versus placebo. Construction Improved Worsened Orientation to name Figure 4: Cognitive function domains Change from baseline in severe impairment battery subscale scores at week 26 (intention-to-treat population). Bars indicate 95% CI for the comparison with baseline. *p= p= p=0 002 versus placebo. * intention-to-treat (ITT) population (all randomly assigned patients who took at least one dose of double-blind trial medication and who had baseline and post-baseline observations). The primary analysis was done on completers (patients in the ITT population who had 6-month data for both primary parameters). A last observation carried forward (LOCF) analysis was done as a sensitivity analysis, and a repeated measurements analysis with an unstructured covariance matrix for the serial correlation was also done. Descriptive statistics were generated for baseline and demographic data. Treatment effects were investigated by means of analysis of covariance (ANCOVA) on the changes at the different timepoints compared with baseline, with factors for treatment, baseline value, baseline MMSE score (continuous covariate), and country. Because there were too many study sites for a useful analysis with site as a covariate, country was used instead (Norway was pooled with Sweden, and the Netherlands with Belgium). Both primary parameters needed to be significant at the 5% significance level; therefore, adjustment for many comparisons was not required. 22 This trial is registered with ClinicalTrials.gov, number NCT Role of the funding source The sponsor was responsible for protocol development, data collection, and data analysis. All authors had full access to the data. The corresponding author had final responsibility for the decision to submit the manuscript for publication. Results Participants were recruited between December, 2003, and March, 2007; the last participant completed doubleblind treatment in September, patients were screened, of whom 407 were randomised to either galantamine or placebo (figure 1). The main reason for screening failure (71 of 98 [72%]) was failure to meet the selection criteria. The patients who were randomised were of advanced age (mean 84 years), and 81% (329 of 407) were women. Demographic and baseline characteristics were well balanced between the two groups (table 1). Comorbidities were present in most patients at baseline; the most common active conditions at screening were cardiovascular, psychiatric, musculoskeletal, and gastrointestinal (table 1). Although more patients in the galantamine group seemed to have psychiatric comorbidities than those in the placebo group, no differences in use of psycholeptic medications were found between the two treatment groups at baseline or during the study (data not shown). 81% (168 of 207) of patients randomly assigned to galantamine and 81% (161 of 200) in the placebo group completed 6 months of treatment. The mean last dose of galantamine was 22 0 (SD 4 2) mg/day and placebo was 22 5 (SD 3 8) mg/day. Of the patients who received 42 Vol 8 January 2009

5 galantamine and completed 6 months of treatment, 146 of 168 (87%) received 24 mg/day, 20 (12%) received 16 mg/day, and two (1%) received 8 mg/day. The mean duration of treatment with galantamine was (SD 44 5) days, and (SD 46 5) days for placebo. Completer analysis of the primary parameters was done on only 311 patients (162 in the galantamine group and 149 in the placebo group) owing to missing observations for these parameters at baseline or week 26 in 18 patients. The mean total SIB score of the galantamine group increased to 69 1 (SD 23 4) points at week 26, which was an improvement of 1 9 (95% CI 0 1 to 3 9) points (figure 2). The mean SIB score in the placebo group decreased by 3 0 ( 5 6 to 0 5) points to 66 9 (SD 23 6). The betweengroup least squares mean difference was 4 36 (95% CI 1 3 to 7 5; p=0 006). There was no significant interaction between treatment effect and baseline MMSE score (p=0 924) or country (p=0 265). The mean total seven-item MDS-ADL self-performance score worsened in both groups: scores at week 26 were 13 0 (SD 7 7) points in the galantamine group (difference at week 26 vs baseline 1 2, 95% CI 0 6 to 1 8) and 13 6 (SD 8 0) points in the placebo group (difference at week 26 vs baseline 1 6, 0 8 to 2 3; figure 3). The between-group least squares mean difference was 0 41 points (95% CI 1 3 to 0 5; p=0 383). 364 patients (186 in the galantamine group and 178 in the placebo group) were included in the LOCF analysis. The mean SIB score in the galantamine group increased to 69 3 (SD 23 2) points at endpoint, which was an improvement of 2 0 (95% CI 0 1 to 3 9) points. In the placebo group, the mean SIB score decreased by 3 2 points (95% CI 5 4 to 0 9) to 65 3 (SD 24 3) points. The between-group least squares mean difference was 5 02 points (95% CI 2 17 to 7 86; p=0 0006). The mean total seven-item MDS-ADL self-performance score in the galantamine group worsened at endpoint to 13 1 (SD 7 7) points in the galantamine group and to 14 0 (SD 8 1) points in the placebo group. Changes from baseline in the seven-item MDS-ADL self-performance score were 1 3 (95% CI 0 7 to 1 9) points and 1 7 (1 0 to 2 4) points, respectively. The between-group least squares mean difference was 0 50 (95% CI 1 39 to 0 39; p=0 394). The results of the repeated measures analysis on both primary parameters were similar to those of the completer and LOCF analyses (SIB: p=0 039 at week 8, p=0 001 at week 12, and p=0 001 at week 26; seven-item MDS-ADL: p=0 965 at week 8, p=0 217 at week 12, and p=0 347 at week 26). Improvements from baseline score were recorded for six of the nine SIB subscales for the patients in the galantamine group compared with a decrease from baseline in all subscales in the placebo group (figure 4). Significant between-group differences were seen in the galantamine group for memory (p=0 006), praxis (p=0 010), and visuospatial ability (p=0 002), with p values Mean change from baseline in MDS-ADL subscale score Bed mobility Transfer (n=166) (n=151) Walk in room Walk in corridor (n=207) Locomotion on unit Locomotion off unit Dressing between 0 05 and 0 10 for language (p=0 064) and attention (p=0 075). Scores for all eleven-item MDS-ADL self-performance subscales worsened in both treatment arms, although the changes were generally smaller numerically in the galantamine group (figure 5). The deterioration in the subscale score for locomotion on unit was significantly less in the galantamine group (p=0 021). During the study, 88% of patients who received galantamine and 89% who received placebo had at least one adverse event (table 2). The investigator deemed these events to be at least possibly related to study medication in 34% of the patients who received galantamine and 28% of the patients who received * (n=200) Any adverse event 183 (88%) 177 (89%) Urinary tract infection 36 (17%) 44 (22%) Vomiting 34 (16%) 30 (15%) Diarrhoea 27 (13%) 37 (19%) Nausea 25 (12%) 13 (7%) Fall 24 (12%) 22 (11%) Agitation 19 (9%) 10 (5%) Arthralgia 15 (7%) 8 (4%) Constipation 14 (7%) 10 (5%) Aggression 14 (7%) 7 (4%) Depression 12 (6%) 6 (3%) Cystitis 11 (5%) 4 (2%) Pyrexia 11 (5%) 10 (5%) Anorexia 11 (5%) 6 (3%) Peripheral oedema 11 (5%) 4 (2%) Bronchitis 8 (4%) 16 (8%) Headache 8 (4%) 13 (7%) Cough 7 (3%) 12 (6%) Hypertension 3 (1%) 11 (6%) Eating Toilet use Personal hygiene Bathing Worsened Figure 5: Specific activities of daily living Change from baseline in eleven-item MDS ADL subscale scores at week 26 (intention-to-treat population). Bars indicate 95% CI for the comparison with baseline. *p=0 021 versus galantamine. Table 2: Treatment-emergent adverse events in more than 5% of patients in either treatment group Vol 8 January

6 Systolic blood pressure (mm Hg) 0 6 ( 3 7 to 2 6) [n=175] 0 3 ( 3 2 to 2 7) [n=161] Diastolic blood pressure (mm Hg) 1 6 ( 3 2 to 0 1) [n=175] 0 8 ( 2 4 to 0 9) [n=161] Heart rate (bpm) 2 7 ( 4 5 to 0 9)* [n=174] 0 9 ( 0 9 to 2 7) [n=161] Bodyweight (kg) 1 5 ( 2 1 to 0 9) [n=164] 0 9 ( 1 5 to 0 3) [n=150] Data are mean (95% CI). *p=0 003 versus screening; p=0 002 between groups; p< versus screening; p=0 002 versus screening. Table 3: Changes in cardiovascular parameters and bodyweight during the study (n=190) (n=185) Heart rate Abnormally low ( 50 bpm) 13 (7%) 8 (4%) Abnormally high ( 120 bpm) 3 (2%) 4 (2%) PQ interval (ms)* Abnormally high ( 210 ms) 14 (8%) 13 (8%) QRS width (ms) Abnormally high ( 120 ms) 10 (5%) 10 (5%) QTcF Borderline (>450 ms) 19 (10%) 21 (11%) Prolonged (>480 ms) 8 (4%) 7 (4%) Pathologically prolonged ( 500 ms) 9 (5%) 9 (5%) Change in QTcF Increase of ms 35 (19%) 26 (14%) Increase of >60 ms 20 (11%) 18 (10%) * n=166; placebo, n=166. n=189. Table 4: Incidence of treatment-emergent ECG abnormalities during the study placebo. The most common adverse events in both treatment groups were urinary tract infections, vomiting, diarrhoea, nausea, and falls (table 2), although most events were mild to moderate. Infections or infestations were reported by 39% of patients in the galantamine group and 44% of patients in the placebo group, with pneumonia or bronchopneumonia in 1% and 4%, respectively. 15% of patients who received galantamine and 16% who received placebo discontinued treatment due to an adverse event. Serious adverse events occurred in 18% of patients who received galantamine and 21% of patients who received placebo; these were deemed to be related to treatment in 3% and 4%, respectively. Deaths occurred in significantly fewer patients who received galantamine compared with those who received placebo (4% vs 11%; p=0 012). Infection-related deaths occurred in 3% of the placebo group, including three deaths that were related to pneumonia or bronchopneumonia. None of the patients who received galantamine died as the result of an infection. Table 3 shows changes in cardiovascular parameters and bodyweight. The incidence of ECG abnormalities was similar in the two groups (table 4). The most commonly reported cardiovascular adverse events were atrial fibrillation (2% in both groups) and bradycardia (1% in galantamine and 2% in placebo). Discussion Here, we report the efficacy of galantamine on cognitive function in patients with severe AD in a randomised, placebo-controlled trial. The results are consistent with those previously reported, and extend the efficacy range from patients with mild to moderate 23,24 and advanced moderate 10,11 into severe AD. The ADAS-cog is widely used to assess cognitive function in clinical trials of patients with mild to moderate dementia, but the scale loses sensitivity in patients with more advanced AD. Therefore, the SIB was developed and validated to assess patients with severe dementia, 18 in which it has shown adequate sensitivity in clinical trials, although these trials were not published when SERAD was initiated. 13,25 The effect of galantamine on cognitive function reported in this study was commensurate with that reported in the nursing home study of donepezil for severe AD. 13 However, direct comparison of the results of these studies was hampered by differences in the severity of cognitive impairment at baseline. By use of covariance analysis of the SERAD data, however, we found no significant interaction between treatment effect and severity of cognitive impairment (baseline MMSE), indicating that there is no reason to assume that the efficacy of galantamine would vary across the range of MMSE values studied. Current guidance from the UK National Institute for Health and Clinical Excellence (NICE) suggests that treatment with inhibitors of cholinesterase should be stopped when the patient s MMSE score reaches 10 points; this decision is based on estimations of cost effectiveness, 26 and the aim of SERAD was not to question the validity of this guideline. Rather than investigate withdrawal in patients who already take galantamine, we investigated the benefit of starting galantamine in patients who do not receive such therapy, despite having advanced cognitive decline. The data show that there are untreated patients with severe AD in nursing homes who could benefit from starting treatment with galantamine. Regardless of whether the same holds for patients with severe AD who are considering continuing with galantamine, the results from a very conservative viewpoint do not support withdrawal of such treatment. There was no statistically significant difference in activities of daily living (MDS-ADL). However, the MDS-ADL was neither developed nor validated to assess differences in ADL scores, but rather to describe disability during a progressing disorder. MDS-ADL was nevertheless selected because, at the time SERAD was designed and initiated, there was no alternative validated scale to assess activities of daily living in patients with severe AD in nursing homes. Similarly, the sensitivity for change with regard to intervention 44 Vol 8 January 2009

7 was unknown because the scale had never been used in a clinical trial, although it was widely used in nursing homes across the USA to assess disability. 20 As a result, the statistical power calculations for SERAD were based on a large US dataset, to which the SERAD inclusion and exclusion criteria were applied. In the donepezil study in patients in nursing homes, which was started just before the SERAD study, the Alzheimer s Disease Cooperative Study activities of daily living inventory for severe AD (ADCS ADL-sev) was used, and the difference between active treatment and placebo was significant. 13 In an earlier study of patients with severe AD (MMSE 1 10) who received donepezil or placebo, no statistically significant difference on ADCS ADL-sev could be found. 14 Analysis of a subgroup of patients with severe AD (MMSE 5 12) in the study of donepezil in moderate to severe AD (MSAD study), however, did significantly favour donepezil over placebo, although a different scale for activities of daily living the disability assessment in dementia scale was used. 27 From the conflicting results of these studies and those of the present study, changes in activities of daily living, even in basic functions, seem to be difficult to detect in patients with severe AD. These mixed effects of treatment on patients functioning are at odds with the effects seen for cognition, for which all three donepezil studies 13,14,27 and the results reported here show a significant beneficial treatment effect over placebo in this severely affected population. The adverse events reported during the study were generally those expected in a population of elderly patients with AD, commonly with multiple comorbidities. Adverse events were usually mild to moderate, with no difference between the treatment groups in those who discontinued treatment as a result of adverse events. The safety of galantamine is also indicated by no difference in cardiovascular and ECG abnormalities between the two groups. Compared with the results of previous trials of galantamine, 23,24 there was an apparent absence of adverse events in the galantamine group relative to the placebo group that might be related to the acetylcholine deficit of patients with severe AD, which is larger than in patients with mild-to-moderate AD. Therefore, after galantamine intake, the peaks and overall increase in acetylcholine might have fewer adverse effects. Very elderly patients might also perceive not only less hunger and thirst but also less nausea. Finally, it should be noted that some risk groups were excluded from the study (as per the exclusion criteria), although this was due to the product labelling and clinical use of galantamine. Generally speaking, few data are available that relate to very old individuals, and the results of SERAD are reassuring for physicians who wish to continue galantamine treatment in patients that enter nursing homes. The finding that significantly fewer deaths occurred in the galantamine group than in the placebo group is particularly noteworthy, and the possible explanations are open to speculation. The overall results suggest that patients in the galantamine group were more active than those who were assigned placebo, with improvements in activities of daily living that related to locomotion. The question of whether treatment influences mortality is, however, complicated. A review of treatment with inhibitors of cholinesterase done by the Swedish Council on Technology Assessment in Health Care Dementia Project concluded that there were no differences in mortality in the randomised trials, 28 with similar findings in a more recent meta-analysis of galantamine trials in mild-to-moderate AD, AD with cerebrovascular disease, or vascular dementia. 29 Conversely, the investigators in a trial of patients with mild cognitive impairment reported higher mortality in the galantamine arm, although overall mortality rates were low. 30 In SERAD, the lower mortality rate in the galantamine arm might have been associated with a reduction in the risk of pneumonia, 31,32 which occurred in more than twice as many patients in the placebo group than in the galantamine group, and was responsible for three deaths in the placebo group. One strength of SERAD is that it recruited elderly patients with severe AD from institutions in several European countries. Nursing homes are known to differ widely between countries, as does the population of resident patients and the skills and training of the staff; these differences were also witnessed during the study. Compared with studies that are done in one country, 13 the inclusion of an international patient population in the SERAD study enables generalisation of the results to nursing home patients with AD across Europe. Indeed, no interaction between country and treatment effect was noted, despite the inclusion of patients from many countries who were treated under various medical practices. The size and age of the patient cohort are also noteworthy, particularly owing to the scarcity of available data on older patients with severe AD; the MMSE range of 5 12 points in a nursing home for patients with AD is representative of most patients with severe AD in Europe. Previous studies of Alzheimer s disease treatments, even in patients with more severe disease, have generally included younger populations (mean age years 10,11,14,27,33 compared with 84 years in the present study). As discussed above, the main restriction of this study was the choice of the MDS-ADL as the scale to assess activities of daily living. This study assessed only the effects of starting galantamine in patients with severe AD, rather than the possible benefits of continuing treatment in patients at this stage. Furthermore, the evaluation of treatment effect in the subgroups of patients with probable AD and possible AD with cerebrovascular disease produced results that were qualitatively similar to those seen in the overall analysis, but that were not significant owing to their inadequate statistical power (data not shown). Vol 8 January

8 In conclusion, data from SERAD show that galantamine can be started and used safely, with good tolerability, in a group of aged patients with severe AD and comorbidities. significantly improved cognitive function, with particular reference to memory, praxis, and visuospatial ability. In addition, there was no significant overall effect on the co-primary parameter of activities of daily living, although there was some evidence of improvement in specific activities of daily living that are related to locomotion. was also associated with a lower mortality rate than was placebo; the reasons for this are unknown, but might be related to the higher physical and mental activity in patients who received galantamine compared with those on placebo. The SERAD investigators group Belgium P P De Deyn, M Vandewoude (Antwerp); J Gillet, J-C Lemper (Brussels); M Gonce (Liège); D Lauwerier (Gavere); E Triau (Leuven); D E J Van De Looverbosch (Deurne); W Van Hooren (Wilrijk); V Verbeeren (Namur). Finland K Halinen, M Hyvärinen, M H Kähönen-Väre (Helsinki); U Parviainen, M M Raivio (Lahti); H Pirttiaho (Oulu); L Seinelä (Tampere); France J Ankri (Paris); F Blanchard (Reims); P Boyer (Chateau de la Valliere and Luynes); X Cnockaert (Beauvais); J-L Duros (Auxerre); A-M Mathieu (Issy les Moulineaux); F Schmitt (Joue les Tours). Italy R Bernabei (Roma); D Cucinotta (Bologna); E Palummeri (Genova); L M Pernigotti (Torino). Netherlands V G M Chel (Amsterdam). Norway D Aarsland, A Øvereng, S Sudkamp (Stavanger); H Helvig (Hundvag); K Mikkelsen (Hafrsfjord); A K Poole, F Skarderud (Oslo). Spain: P Abizanda (Albacete); J Álaba, E Arriola (San Sebastian); M Boada (Barcelona); A J Cruz-Jentoft (Madrid); J Gutiérrez (Aviles). Sweden A Adjemian (Västerås); E Andersson (Halmstad); N Andreasen (Huddinge); S Bergman (Västervik); A Börjesson (Ängelholm); G Dey (Haninge); B Klanger (Västerås); A Larsson (Trollhättan); B Lorentzon (Trollhättan); J Louhija (Uppsala); H Malcus (Göteborg); T Neumark (Kalmar); M Skoglund (Skellefteå); A Wimo (Bergsjö). Switzerland S Bertoli-Edler (Basel); C Hock, M Kirsten-Krüger (Zuerich). United Kingdom R Bullock (Swindon); T Elliott (Shrewsbury); S W Simpson (Dorset); D G Wilkinson (Southampton). Study safety board A Burns (Manchester); L Frölich (Mannheim). Local Trial Management for Janssen Cilag. Belgium E De Schrijver, E Present, L Teugels, B Vanneste, C Wouters. Finland J Hemdahl, K Kettunen, L Rahkamo, M Stenstrand, M Utriainen. France M G Commenil, A Delgado, D Guitton, P Motlagh, F Nave, J Pinard. Italy M Adami, V Carrera. Netherlands E Delarue, E De Rijck, S Gurdebeke, S S Vanderplaetsen. Norway I Robertsen, I Salvesen, O-C Skare, B Varaas. Spain C Bas, J M Blasco, C Guisado, C Guzmán, P Martínez, C Mendoza. Sweden B Eriksson, M Rehmar, M Spinelli Scala. Switzerland E Hroarsdottir, M Rimle. Local trial management for Shire. United Kingdom H Long, J Wallis, E Zuccaro-Boycott, I Howe. Central trial management. E Came, D De Wulf, G Hammond, E Kitto, J C van Oene. Study responsible physician S Schwalen. Data management J Beukelaers, I Vandriessche, S Van Genechten. Analysis and statistics B Van Baelen. Contributors AB and LF were members of the steering and safety-monitoring committee for the SERAD study. RBe, RBu, AJCJ, CH, MR, ET, MV, and AW participated in the SERAD study as investigators. EC was responsible for the central management of the trial, the review of the study report, early discussion of manuscript content, and review of draft manuscripts. BVB did statistical analysis of the data. GLH helped to design the study and project managed it from 2003 to JCvO participated in the supervision of central trial management, which involved trial conduct, data management, and analysis. SS assisted with trial design and analysis of data and participated as an investigating physician. All authors participated in drafting and approved the final manuscript. Conflicts of interest AB receives research funding and consultancy fees from pharmaceutical companies involved in the manufacture and marketing of drugs for Alzheimer s disease, including Janssen Cilag, Pfizer, Novartis, Eisai, Shire, and Lundbeck. RB receives consultation fees from Boehringer Ingelheim and Angelini Farmaceutici, and research grants from MSD, Angelini, and the European Union Commission. RB has participated in clinical trials, given presentations, and sat on advisory boards for Johnson & Johnson. ACJ has received consultancy fees and funding for educational activities from pharmaceutical companies involved in the manufacture and marketing of drugs for Alzheimer s disease, including Janssen Cilag, Novartis, Esteve, Eisai, and Lundbeck. LF receives research funding and consultancy fees from pharmaceutical companies involved in the manufacture and marketing of drugs for Alzheimer s disease, including AstraZeneca, Janssen Cilag, Pfizer, Merz Pharma, Myriad Pharmaceuticals, Elan/Wyeth, Novartis, Eisai, and Lundbeck. CH receives research funding or consultancy fees from several pharmaceutical, biotech, or venture capital companies involved in the R&D, manufacturing, marketing, or financing of (potential) drugs for Alzheimer s disease: Janssen Cilag, Pfizer, Novartis, Roche, GSK, MSD, Merz, Lundbeck, Evotec, Neurimmune, TVM Capital, Novartis Venture Funds, Nextech Venture, and Debiopharm. In addition, he is a scientific founder, shareholder, and board member of Neurimmune Therapeutics AG, Zurich, Switzerland. MR has cooperated in clinical trials, received research and lecture fees and grants to attend international congresses from pharmaceutical companies involved in the manufacture and marketing of drugs for Alzheimer s disease, including Janssen Cilag, Pfizer, Novartis, and Lundbeck. ET receives research funding and consultancy fees from pharmaceutical companies involved in the manufacture and marketing of drugs for Alzheimer s disease, including Janssen Cilag, Pfizer, Novartis and Lundbeck. MV receives research funding and consultancy fees from pharmaceutical companies involved in the manufacture and marketing of drugs for Alzheimer s disease, including Janssen Cilag, Pfizer, and Lundbeck. AW acted as consultant to drug companies that are purchasing or developing drugs for treatment of Alzheimer s disease or other dementias (Pfizer, Janssen Cilag, Novartis, Merz, Lundbeck, Forest, GSK, Wyeth, Sanofi, Elan, Neurochem). EC is a freelance clinical research consultant and was contracted by Janssen Cilag to provide central study management. BvB is an independent consultant working for SGS Life Sciences Services who was hired by Janssen Cilag to undertake statistical analysis. GH is an employee of Janssen Cilag Ltd, Global Clinical Operations Division. JCvO is an employee of Janssen Cilag EMEA (Europe, Middle East and Africa). SS is a former employee of Janssen Cilag EMEA (Europe, Middle East, and Africa). Acknowledgments This study was funded by Janssen-Cilag EMEA. Assistance with the protocol development was provided by S Kavanagh (Jansen Cilag PGSM, Beerse, Belgium). Editorial assistance with drafting and completion of the manuscript was provided by D Booth (Bioscript Stirling Ltd, London, UK), who is a medical writer paid by Janssen-Cilag EMEA. References 1 Wancata J, Musalek M, Alexandrowicz R, Krautgartner M. Number of dementia sufferers in Europe between the years 2000 and Eur Psychiatry 2003; 18: Desai AK, Grossberg GT, Sheth DN. Activities of daily living in patients with dementia: clinical relevance, methods of assessment and effects of treatment. CNS Drugs 2004; 18: Yaffe K, Fox P, Newcomer R, et al. Patient and caregiver characteristics and nursing home placement in patients with dementia. JAMA 2002; 287: Jönsson L, Eriksdotter Jönhagen M, Kilander L, et al. Determinants of costs of care for patients with Alzheimer s disease. Int J Geriatr Psychiatry 2006; 21: Maelicke A, Albuquerque EX. Allosteric modulation of nicotinic acetylcholine receptors as a treatment strategy for Alzheimer s disease. Eur J Pharmacol 2000; 393: Aguero-Torres H, Fratiglioni L, Winblad B. Natural history of Alzheimer s disease and other dementias: review of the literature in the light of the findings from the Kungsholmen Project. Int J Geriatr Psychiatry 1998; 13: Vol 8 January 2009

9 7 Loy C, Schneider L. for Alzheimer s disease and mild cognitive impairment. Cochrane Database Syst Rev 2006: CD Loveman E, Green C, Kirby J, et al. The clinical and costeffectiveness of donepezil, rivastigmine, galantamine and memantine for Alzheimer s disease. Health Technol Assess 2006; 10: Folstein MF, Folstein SE, McHugh PR. Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: Wilkinson DG, Hock C, Farlow M, van Baelen B, Schwalen S. provides broad benefits in patients with advanced moderate Alzheimer s disease (MMSE < or = 12) for up to 6 months. Int J Clin Pract 2002; 56: Blesa R, Davidson M, Kurz A, et al. provides sustained benefits in patients with advanced moderate Alzheimer s disease for at least 12 months. Dement Geriatr Cogn Disord 2003; 15: Feldman H, Gauthier S, Hecker J, et al. A 24-week, randomized, double-blind study of donepezil in moderate to severe Alzheimer s disease. Neurology 2001; 57: Winblad B, Kilander L, Eriksson S, et al. Donepezil in patients with severe Alzheimer s disease: double-blind, parallel-group, placebocontrolled study. Lancet 2006; 367: Black SE, Doody R, Li H, et al. Donepezil preserves cognition and global function in patients with severe Alzheimer disease. Neurology 2007; 69: American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th edn. Washington, DC: American Psychiatric Association, McKhann G, Drachman D, Folstein M, et al. Clinical diagnosis of Alzheimer s disease: report of the NINCDS ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer s Disease. Neurology 1984; 34: Roman GC, Tatemichi TK, Erkinjuntti T, et al. Vascular dementia: diagnostic criteria for research studies. Report of the NINDS AIREN International Workshop. Neurology 1993; 43: Albert M, Cohen C. The test for severe impairment: an instrument for the assessment of patients with severe cognitive dysfunction. J Am Geriatr Soc 1992; 40: Hawes C, Morris JN, Phillips CD, et al. Development of the nursing home Resident Assessment Instrument in the USA. Age Ageing 1997; 26 (suppl 2): Morris JN, Fries BE, Morris SA. Scaling ADLs within the MDS. J Gerontol A Biol Sci Med Sci 1999; 54: M Panisset M, Roudier M, Saxton J, Boller F. Severe impairment battery. A neuropsychological test for severely demented patients. Arch Neurol 1994; 51: ICH. Guidance on statistical principles for clinical trials Available at: (accessed Nov 18, 2008). 23 Raskind MA, Peskind ER, Wessel T, Yuan W. in AD: a 6-month randomized, placebo-controlled trial with a 6-month extension. The USA-1 Study Group. Neurology 2000; 54: Tariot PN, Solomon PR, Morris JC, et al. A 5-month, randomized, placebo-controlled trial of galantamine in AD. The USA-10 Study Group. Neurology 2000; 54: Reisberg B, Doody R, Stoffler A, et al. Memantine in moderate-tosevere Alzheimer s disease. N Engl J Med 2003; 348: UK National Institute for Health and Clinical Excellence. Donepezil, galantamine, rivastigmine (review) and memantine for the treatment of Alzheimer s disease. London: National Institute for Health and Clinical Excellence, Feldman H, Gauthier S, Hecker J, et al. Efficacy and safety of donepezil in patients with more severe Alzheimer s disease: a subgroup analysis from a randomized, placebo-controlled trial. Int J Geriatr Psychiatry 2005; 20: SBU. Dementia and mortality. In: Dementia: a systematic review, vol 2. Stockholm: The Swedish Council on Technology Assessment in Health Care (SBU), 2008: Feldman HH, Pirttila T, Dartigues JF, et al. Analyses of mortality risk in patients with dementia treated with galantamine. Acta Neurol Scand 2008; published online April 3. DOI: / j Winblad B, Gauthier S, Scinto L, et al. Safety and efficacy of galantamine in subjects with mild cognitive impairment. Neurology 2008; 70: Marrie TJ. Community-acquired pneumonia in the elderly. Clin Infect Dis 2000; 31: Riquelme R, Torres A, El-Ebiary M, et al. Community-acquired pneumonia in the elderly: a multivariate analysis of risk and prognostic factors. Am J Respir Crit Care Med 1996; 154: Feldman HH, Schmitt FA, Olin JT. Activities of daily living in moderate-to-severe Alzheimer disease: an analysis of the treatment effects of memantine in patients receiving stable donepezil treatment. Alzheimer Dis Assoc Disord 2006; 20: Vol 8 January