Protocol. This trial protocol has been provided by the authors to give readers additional information about their work.

Protocol This trial protocol has been provided by the authors to give readers additional information about their work. Protocol for: Howard R, McShane R, Lindesay J, et al. Donepezil and memantine for moderate-to-severe Alzheimer s disease. N Engl J Med 2012;366:893-903.

DOM NO-AD Donepezil and Memantine in Moderate to Severe Alzheimer s Disease Donepezil and Memantine in Moderate to Severe Alzheimer s disease Final Version 6.1 16/06/2010 Trial Identifier: EUDRACT Number: 2007-001172-36 Protocol Authorisation: Chief Investigator: Professor Robert Howard Signature:.. Date: Chair of Trial Steering Committee: Signature:.. Date: Chair of Data Monitoring Committee: Signature:.. Date:

CONTENTS 1 GENERAL INFORMATION 6 Protocol Information 6 1.1.1 Compliance 6 1.1.2 Peer-review 6 1.2 Main Contacts 6 1.2.1 Sponsor 6 1.2.2 Chief Investigator 6 1.2.3 Trial Manager 6 1.2.4 Randomisation Centre 7 1.2.5 Medication Supply 7 1.2.6 Central Packaging & Labelling and Distribution to Local Pharmacies 7 1.3 Trial Committees 7 1.3.1 Trial Steering Committee 7 1.3.2 Independent Data Monitoring Committee 8 1.3.3 Trial Management Group 8 1.4 Declarations of Competing Interests 9 1.4.1 Potential Competing Interests for Investigators 9 1.4.2 No Competing Interests for Investigators 9 1.4.3 Competing Interests for Investigators 10 2 ABBREVIATIONS 12 3 SUMMARY 14 3.1 Synopsis 14 3.1.1 Type of Design 14 3.1.2 Patients Studied 14 3.1.3 Objectives 14 3.1.4 Outcome Measures 15 3.1.5 Trial Interventions 15 3.1.6 Planned inclusion/exclusion criteria 16 3.1.7 Duration 16 3.1.8 Risks and Benefits 16 3.1.9 Organisation 16 4 BACKGROUND INFORMATION 17 4.1 Introduction 17 4.2 Clinical Studies 18 4.3 National Institute of Clinical Excellence Guidance 19 4.4 Choice of Study Population 20 4.5 Choice of Investigational Drugs 20 4.6 Choice of Outcome measures 20 4.7 Consumer Involvement 21 4.8 Risks and Benefits 21 4.8.1 Potential Risks 21 4.8.2 Potential Benefits 21 5 TRIAL OBJECTIVE AND PURPOSE 22 Page 2 of 68

5.1 Aim 22 5.2 Objectives 22 5.2.1 Primary Objectives 22 5.2.2 Secondary Objectives 22 6 TRIAL DESIGN 24 6.1 Design 24 6.2 Trial Flow Charts 24 6.3 Duration 25 6.3.1 Duration of the treatment period 25 6.3.2 Duration of the follow-up period 25 6.3.3 Definition of completion of the trial for an 25 individual participant 6.3.4 Definition of the end of the trial 25 6.4 Premature Trial Closure 25 7 SELECTION AND WITHDRAWAL OF SUBJECTS 26 7.1 Number and Source of Participants 26 7.2 Planned recruitment rate 26 7.3 Recruitment Strategy 27 7.4 Monitoring and ensuring recruitment to trial 27 7.5 Consent Procedure 27 7.6 Eligibility Criteria 28 7.6.1 Inclusion Criteria 28 7.6.2 Exclusion Criteria 28 7.6.3 Contraindications, cautions and interactions 29 7.7 Randomisation and Enrolment Procedure 29 7.7.1 Method of identification of participants and carers 29 7.7.2 Randomisation 29 7.8 Assessment Schedule 30 7.9 Patient Discontinuation 31 7.9.1 Discontinuation from treatment only 31 7.9.2 Discontinuation from treatment and follow up 31 7.10 Patient transfers 31 7.11 Loss to follow up 31 8 TREATMENT REGIMEN 32 8.1 Planned Trial Interventions 32 8.2 Description of randomised procedures 33 8.2.1 Donepezil and Memantine 33 8.2.2 Placebo 33 8.2.3 Double dummy design 33 8.2.3.1Titration schedule 33 8.2.3.2 Maintenance schedule 33 8.3 Dosing 34 8.4 Blinding 34 8.5 Packaging and labelling of Investigational Medicinal 34 Product 8.6 Distribution of Investigational Medicinal Product 34 8.7 Prescription of Investigational Medicinal Product 34 Page 3 of 68

8.8 Administration of Investigational Medicinal Product 35 8.9 Unused Trial Medication and Study Medication 35 Accountability 8.10 Concomitant Medication 35 8.11 Departures from Randomised Treatment 35 8.11.1 Compliance 35 8.11.2 Emergency unblinding procedures 35 8.11.3 Treatment at the end of trial 36 9 ASSESSMENT OF EFFICACY 37 9.1 Primary Measures 37 9.1.1 Standardised Mini Mental State Exam 37 9.1.2 Bristol Activities of Daily Living Scale 37 9.2 Secondary Measures 37 9.2.1 Client Service Receipt Inventory 37 9.2.2 EuroQol EQ-5D 37 9.2.3 Neuropsychiatric Inventory 37 9.2.4 DEMQOL proxy 38 9.2.5 General Health Questionnaire 12 38 9.2.6 Institutionalisation 38 10 ASSESSMENT OF SAFETY 39 10.1 Adverse Events and Adverse Reactions 39 10.2 Serious Adverse Events 39 10.3 Serious Adverse Reactions 39 10.4 Suspected Unexpected Serious Adverse Reactions 39 10.5 Assessment of Causality and Severity 39 10.5.1 Causality 39 10.5.2 Seriousness 40 10.5.3 Expectedness 40 10.6 Reporting 41 10.6.1 Reporting responsibilities 41 11 STATISTICS 44 11.1 Statistical Considerations 44 11.2 Health Economic Evaluation 44 11.3 Independent data monitoring and ethics committee: determining when clear answers have emerged 45 12 DATA MANAGEMENT AND MONITORING 46 12.1 Direct Access to Source Data and Documents 46 12.2 Confidentiality 46 12.3 Record Keeping 46 12.3.1 Custodian of the data 46 12.3.2 Format of records 46 12.3.3 Duration and location 46 12.4 Trial Data Management System 47 12.4.1 ecrf 47 12.4.2 Training and User Support 47 12.4.3 Entry of Data by Research Workers 47 Page 4 of 68

13 QUALITY CONTROL AND ASSURANCE 49 13.1 Quality Assurance 49 13.1.1 Training of trial personnel 49 13.1.2 Essential Documentation 49 13.1.3 Standard Operating Procedures 49 13.1.4 Risk and Quality Management Plan 49 13.2 Quality Control 49 13.2.1 On-site monitoring 49 13.2.2 Data checking and verification procedures 49 13.2.3 Data Lock 50 14 ETHICAL CONSIDERATIONS 51 15 REGULATORY APPROVAL 52 15.1 Research Ethics Approval 52 15.2 Site Specific Information Form 52 15.3 Medicines and Healthcare Products Regulatory Agency 52 Approval 16 FINANCING AND INSURANCE 53 16.1 Funding arrangements 53 16.1.1 Contact details of funding bodies 53 16.1.2 Duration of grant 53 16.1.3 Grant summary 53 16.2 Indemnity 53 16.3 Site Agreements 53 17 PUBLICATION POLICY 54 18 ANCILLARY STUDIES 55 19 AMENDMENTS 56 20 REFERENCES 59 21 APPENDICES 61 Appendix 1: Further contact details Appendix 2: Summary of medical characteristics for donepezil and memantine Appendix 3: Donepezil and memantine dosing schedule Page 5 of 68

1 GENERAL INFORMATION 1.1 Protocol Information 1.1.1 Compliance The trial will be conducted in compliance with the protocol, the European Union Clinical Trials Directive (2001/20/EC), the associated UK Medicines for Human Use (Clinical Trials) Regulations (2004) and Medicines for Human Use (Clinical Trials) Amendment Regulations 2006, the Data Protection Act (1998), Ethics Committee and MHRA approvals, the principles of ICH Good Clinical Practice (GCP) guidelines (CPMP/ICH/135/95), the principles of the Declaration of Helsinki (1996) and other requirements as appropriate. 1.1.2 Peer-Review This study has been subject to intensive independent anonymous peer review by the Medical Research Council prior to their making their decision to fund this study. 1.2 Main Contacts 1.2.1 Sponsor Name: Kings College London, Joint Clinical Trials Office Address: Joint Clinical Trials Office, 16 th Floor, Guy s Tower, London SE1 9RT Telephone: 0207 188 5732 Fax: 0207 1888330 Email: jackie.pullen@kcl.ac.uk 1.2.2 Chief Investigator Professor Robert Howard Professor of Old Age Psychiatry and Psychopathology PO70 Institute of Psychiatry De Crespigny Park London SE5 8AF Tel: 020 7848 0545 Fax: 020 7848 0632 robert.j.howard@kcl.ac.uk 1.2.3 Trial Manager Mary Griffin PO70 Institute of Psychiatry De Crespigny Park London SE5 8AF Tel: 020 7848 0024 Fax: 020 7848 0132 mary.griffin@kcl.ac.uk Page 6 of 68

1.2.4 Randomisation Centre Medical Research Council: Clinical Trials Unit 222 Euston Road London NW1 2DA 1.2.5 Medication Supply Manufacture of Donepezil and matching placebo Eisai Ltd 3, Shortlands London W6 8EE UK Manufacture of Memantine and matching placebo Lundbeck Pharmaceuticals Ottilliavej 9 DK 2500 Valby Denmark 1.2.6 Central Packaging & Labelling and Distribution to Local Pharmacies Karl Jones Catalent Pharma Solutions Clinical Supply Services Wingates Industrial Park Westhoughton Bolton Lancs, BL5 3XX Tel: 01942 790000 Fax: 01942 799799. 1.3 Trial Committees 1.3.1 Trial Steering Committee The Trial Steering Committee (TSC) is responsible for the independent oversight of the progress of the trial, investigation of serious adverse events, and determining the future progress of the trial in light of regular reports from the DMEC and TMG. The TSC has the power to prematurely close the trial. The TSC will meet annually or more often if the chair determines a reason for doing so and is composed of: Professor Cornelius Katona (Chair) Ken Wilson (Independent Old Age Psychiatrist ) Robert Hills (Independent Statistician) Professor Robert Howard (Chief Investigator) Tony Johnson (Statistician) Page 7 of 68

Invited observers: Representative from the funder, the MRC Representative from the sponsor, JCTO Patient / carer representatives from the Alzheimers Society Patrick Phillips (Trial Statistician) Mary Griffin (Trial Manager, Secretary to the TSC) 1.3.2 Independent Data Monitoring Committee The Independent Data Monitoring Committee (IDMC) is independent from the trial and is responsible for monitoring the progress of the trial including: recruitment, protocol adherence, serious adverse events and side effects of treatment as well as the difference between the trial treatments on the primary outcome measures. The IDMC will meet annually or more often if the chair determines a reason for doing so. They will provide a confidential trial progress report at the end of each meeting which will be sent to the TSC. The IDMC will agree their structure and organisation in an IDMC Charter (DAMOCLES Study Group, 2005) before randomisation commences. The DMEC can recommend premature closure of the trial to the TSC in accordance with the IDMC charter. The IDMC is composed of: Brian Lawlor (Chair) Tony Bayer (Independent Physician) Deborah Ashby (Independent Statistician) The trial statistician will be in attendance but not a member 1.3.3 Trial Management Group (TMG) This group will oversee the running of the trial, monitor and maintain recruitment rates, and advise participants on any necessary changes in the running of the trial or analysis of the data. It will also resolve any problems that arise in patient management or conduct of the trial. The full TMG will meet quarterly in the first year and biannually thereafter. It is composed of: Professor Robert Howard (Chair) All Investigators Trial statisticians Health economists Trial manager (Secretary to the TMG) Data manager Sub-committees may be formed from the full TMG for specific purposes (e.g. protocol development, writing papers etc). These committees will be appointed by the full TMG and will meet as necessary. 1.4 Declarations of Competing Interests Many of the investigators have received support from pharmaceutical companies for example to attend conferences, for giving lectures, for the provision of consultancy, or for the conduct of research. A list of investigators competing interests is contained in section 1.4.3. 1.4.1 Potential Competing Interests for Investigators Page 8 of 68

The protocol of the trial states that a central register of competing interests will be held by the trial team of all investigators and colleagues are asked to inform the trial management team if any of this information changes during the course of the trial. Examples of potential competing interests are: a) Stock ownership in any commercial companies involved; b) Stock transaction in any commercial company involved (if previously holding stock); c) Ongoing advisory role to a company providing drugs to the trial; d) Frequent speaking engagements on behalf of one of the interventions; e) Intellectual conflict e.g. strong prior belief in one or more of the trial s experimental arms; f) Involvement in regulatory issues relevant to the trial s procedures; g) Investment (financial or intellectual) or career tied up in competing products. For publication purposes, the details of competing interests for the DOMINO- AD trial are as follows: 1.4.2 No Competing Interests for Investigators Professor Robert Howard, London Maudsley Dr Tom Dening, Cambridge Dr Alan Hughes, Glasgow Professor James Lindesay, Leicester Professor Ian McKeith, Newcastle Dr Rob Jones, Nottingham Professor Clive Holmes, Southampton Dr Bart Sheehan, Warwick Ashley Baldwin, 5 Boroughs John O Brien, Newcastle Bob Barber, Newcastle Robin Jacoby, Oxford Dr Rupert McShane, Oxford Dr Heinrich Lamprecht, Newcastle Dr Paul Koranteng, Northamptonshire Dr Stephen Pearson, Plymouth Dr David Findlay, Dundee Jill Mann, Bath Dr Craig Ritchie, London Imperial Page 9 of 68

1.4.3 Competing Interests for Investigators Professor Sube Banerjee, London Maudsley declared: Development of the DEMQOL system; Received speaker and consultancy fees from all companies involved in making anti-dementia medication and have had an educational grant from Pfizer; Has worked for DH. Dr Peter Passmore, Belfast declared: Has been a member of speaker bureaus in international meetings and conferences for Eisai, Janssen-Cilag, Lundbeck, Novartis, Pfizer and national meetings in UK for Bayer, BMS, Astra Zeneca, Sanofi- Synthelabo, MSD, Eisai, Shire, Norvartis, Pfizer; Has been an advisor internationally for Bayer, BMS, Eisai, Janssen- Cilag, Norvartis, Pfizer, Sanofi-Synthelabo, and nationally in the UK for Bayer, BMS, Astra Zeneca, Sanofi-Synthelabo, MSD, Eisai, Shire, Lundbeck, Norvartis, Pfizer. Professor Roy Jones, Bath declared: Has received grant support, consulting fees and honoraria from companies with products in the field of Alzheimer s disease and related conditions; RICE has recently completed the development of a new research building which has been funded as a result of donations to a major capital appeal. A number of pharmaceutical companies including Lundbeck, Merz, Eisai and Pfizer working in the field of Alzheimer s disease and dementia have contributed to this appeal. Professor Alistair Burns, Manchester declared: Received research funding and honoraria and expenses for consultancy work from companies involved in the manufacturing and marketing of drugs for dementia Eisai, Pfizer, Shire, Baxter, Janssen Does occasional lectures for companies hosting meetings on behalf of these industries Pharam-Ed and Phase-V; Gets paid expenses from the Alzheimer s Society in the UK and Alzheimer s Australia (for lectures in 2008); Received an honorarium from John Wiley for my role as editor of the International Journal of Geriatric Psychiatry and receive honoraria from a number of publishers for books written and edited. Clive Ballard, King s College declared: Received honoraria from Novartis, Pfizer, Shire, Lundbeck, Myriad, Janssen, Astra Zeneca and Servier pharmaceutical companies and Page 10 of 68

research grants from Novartis, Lundbeck, Astra-Zeneca and Janssen pharmaceuticals. Dr Peter Bentham, Birmingham declared: Is a paid consultant to Tau Therapeutics Pte Ltd. Page 11 of 68

2 ABBREVIATIONS AD Alzheimer s Disease ADCS - Alzheimer s Disease Co-operative Study Activities of Daily ADL Living Inventory AE Adverse Event AR Adverse Reaction BADLS Bristol Activities of Daily Living BPSD Behavioural and Psychological Symptoms of Dementia CI Chief Investigator CIBIC Clinician Interview Based Impression of Change CIN Carer Identification Number CSRI Client Service Receipt Inventory CTA Clinical Trial Authorisation CTU Clinical Trials Unit CV Curriculum Vitae DeNDRoN Dementias and Neurodegenerative Diseases Research Network DM Data Manager ecrf Electronic Case Report Form EU European Union EQ-5D EuroQol 5D GCP Good Clinical Practice GHQ General Health Questionnaire GMP Good Manufacturing Practice GP General Practitioner HRQoL Health Related Quality of Living ICH International Conference on Harmonisation IDMC Independent Data Monitoring Committee KCL King s College London LOCF Last Observation Carried Forward MHRA Medicines and Healthcare products Regulatory Authority MHRN Mental Health Research Network MMRM Multilevel modelling repeated measures MRC Medical Research Centre MREC Multi centre Research Ethics Committee NCCHTA National Co-ordinating Centre for Health Technology Assessment NHS National Health Service NICE National Institute for Clinical Excellence NINCDS- National Institute of Neurological and Communicative Diseases ADRDA and Stroke/Alzheimer's Disease and Related Disorders Association NMDA N-methyl-D-aspartic acid NPI Neuropsychiatric Inventory PI Principal Investigator PIN Patient Identification Number QALY Quality adjusted life year QP Qualified Person Page 12 of 68

QRD R&D REC RW SAE SAR s.d. SDW SMMSE SmPC SOPs SSA SUSAR TM TMG TSC UAR UK WIHRD Quality research in Dementia Research and Development Research Ethics Committee Research Worker Serious Adverse Event Serious Adverse Reaction Standard Deviation Source Data Worksheets Standardised Mini Mental State Exam Summary of Product Characteristics Standard Operating Procedures Site Specific Assessment Suspected Unexpected Serious Adverse Reaction Trial Manager Trial Management Group Trial Steering Committee Unexpected Adverse Reaction United Kingdom Wessex Institute for Health Research and Development Table 1: List of abbreviations used in the trial Page 13 of 68

3 SUMMARY 3.1 Synopsis Currently the evidence is unclear on what clinicians should prescribe to patients with Alzheimer s disease once they reach the moderate to severe transition point. The trial will look at whether pharmacological treatment with donepezil alone or memantine alone is better than placebo at maintaining cognitive function and activities of daily living in the patients described above. It will also consider whether there are synergistic effects of the two drugs combined, compared with each drug alone 3.1.1 Type of design This will be a pragmatic, multi-centre, double-blind, randomised, placebocontrolled (double dummy), parallel group, 2X2 factorial clinical trial. 3.1.2 Patients studied Patients recruited to this trial will be diagnosed with Alzheimer s disease which has reached a moderate to severe diagnosis (SMMSE score 5 13) and who are currently taking donepezil. 3.1.3 Objectives Primary Objectives The trial will test a number of hypotheses in memory clinic patients who have declined in terms of cognitive function to reach the transition point to moderate-to-severe AD. a) Patients with AD who continue donepezil beyond the moderate to severe transition point will show a significantly smaller decline on ratings of cognitive function and activities of daily living over the following 12 months than those discontinuing donepezil. b) Patients with AD who commence memantine therapy at the moderate to severe transition point will show a significantly smaller decline on ratings of cognitive function and activities of daily living over the following 12 months than those who do not. c) Patients given the combination of memantine and donepezil at the moderate to severe transition point will show additive or synergistic significant benefits on measures of activities of daily living and cognitive function after 12 months compared to those patients continuing on either monotherapy. Secondary Objectives a) Patients with AD who continue donepezil beyond the moderate to severe transition point will show a significantly smaller deterioration on ratings of non-cognitive symptoms and health related quality of life over the following 12 months than those discontinuing donepezil. Analysis at the margins using 100% of trial participants. Page 14 of 68

b) Patients with AD who commence memantine therapy at the moderate to severe transition point will show a significantly smaller deterioration on ratings of non-cognitive symptoms and health related quality of life over the following 12 months than those who do not. c) Patients given the combination of memantine and donepezil at the moderate to severe transition point will show additive or synergistic significant benefits on measures of non-cognitive symptoms and health related quality of life after 12 months compared to those patients continuing on either monotherapy. d) Treatment of patients with donepezil beyond the moderate to severe transition point will be more cost-effective than discontinuing donepezil. Memantine therapy will be more cost-effective than placebo. The combination of memantine and donepezil will be more cost-effective than monotherapy. e) Patients who continue on donepezil beyond the moderate to severe transition point will be institutionalised later than those who do not. Patients who commence memantine therapy will be institutionalised later than those taking placebo. Patients who commence combination of memantine and donepezil will be institutionalised later than those on monotherapy. For carers, parallel secondary objectives will be changes in psychological morbidity and health related quality life. 3.1.4 Outcome measures The primary outcome measures will be: Cognitive function (measured using the SMMSE) Activities of daily living (measured using the BADLS) The secondary outcome measures will be: Non cognitive dementia symptoms (measured using the NPI) Health related quality of life (measured using EQ-5D and DEMQOLproxy) Care giver burden (measured using the GHQ-12) Cost effectiveness (assessed through consideration of the combination of the CSRI and assessments of function and quality of life) Institutionalisation (defined as permanent transition to a care home, continuing care unit or hospital) 3.1.5 Trial interventions The trial will consist of 4 arms with equal allocation to each: Arm 1) 10mg donepezil plus 20mg memantine Arm 2) Donepezil placebo plus 20mg memantine Arm 3) 10mg donepezil plus memantine placebo Arm 4) Donepezil placebo plus memantine placebo Page 15 of 68

3.1.6 Planned inclusion/exclusion criteria. Inclusion criteria: People will be eligible to participate if they meet standardized clinical (NINCDS-ADRDA) criteria for probable or possible Alzheimer s disease, have been continuously prescribed donepezil for at least 3 months and continuously prescribed 10mg donepezil in previous 6 weeks. They will have had no changes in prescription of drugs for other psychiatric symptoms (e.g. antipsychotic, antidepressant, benzodiazepine) in previous 6 weeks and have deteriorated to the point where there is uncertainty that the person receiving benefit from donepezil and that on testing with a standardized assessment of cognitive function the MMSE score is between 5 and 13. Also to be eligible, the participant must be community based with a resident family or professional carer or visited on at least a daily basis by a carer. The must have given consent if considered capable and the main carer (informal or institutional) must also give consent to their own involvement and to the participant s involvement. Exclusion criteria: These will include severe, unstable or poorly controlled medical conditions, current prescription of memantine, contra-indications or previous adverse or allergic reactions to trial drugs, involvement in another trial or concerns over the patient s compliance. 3.1.7 Duration The proposed duration of the treatment period is 52 weeks post randomisation. All study measures will be assessed at randomisation, at 6 weeks (except DEMQOL-proxy) to address the acute effects of withdrawal of donepezil, and at 18 weeks (except CSRI and GHQ-12), 30 weeks (except DEMQOL-proxy) and at 52 weeks. Participants will then be followed up every 26 weeks for 3 years by telephone interview to establish whether and on what date they have entered a care institution. 3.1.8 Risks and benefits. As both donepezil and memantine are well tolerated drugs the main risk for participants is withdrawal of their existing prescription of donepezil. There is also a risk of minor distress or inconvenience. This is balanced by the fact that although neither donepezil nor memantine are recommended by NICE for the treatment of patients with low MMSE scores on the basis of cost effectiveness, both drugs have shown advantages over placebo on important clinical outcomes in this patient group. 3.1.9 Organisation The trial is funded by the Medical Research Council and the Alzheimer s Society and sponsored by King s College London. Page 16 of 68

4 BACKGROUND INFORMATION 4.1 Introduction Acetylcholinesterase inhibitors are widely prescribed to patients with mild to moderate AD with studies showing they improve cognition and stabilize cognition, function and behaviour for up to 6-12 months. In 2002, the number of GP prescriptions for each of the cholinesterase inhibitors in England was 148,400 for donepezil, 26,100 for galantamine and 27,300 for rivastigmine (Hansard 2003). Although there is evidence that these drugs are effective in the SMMSE 10-24 range, there is little evidence that they are ineffective at SMMSE scores of less than 12 or 10 points. A straw poll of experienced colleagues carried out in 2004 indicated that each had developed their own stopping rules. Some physicians would try a drug holiday after 12 months and only restart treatment if there was evidence of abrupt decline during the withdrawal period. Others said they would stop treatment if they felt that the drug was no longer working judging that the patient was declining as rapidly as might be expected without treatment. Finally, there was a group of physicians who said that they would continue to prescribe indefinitely, regardless of dementia severity or the rate of disease progression. There is at present no evidence base to guide the behaviour of physicians at this critical decision point. The only drug licensed for the treatment of moderate to severe AD in the UK is memantine. In a Cochrane systematic review and meta-analysis, statistically significant results in favour of treatment with memantine 20mg/day in patients with moderate to severe AD were seen on cognition, activities of daily living and global clinical impression of change at 28 weeks (Areosa 2005). Again, however, there is an inadequate evidence base to guide the behaviour of physicians whose patients are compliant with a cholinesterase inhibitor but have reached the moderate to severe point where they might be considered for memantine treatment. There is therefore a pressing clinical need to provide an evidence base for physicians on which to base decisions about continued prescribing of cholinesterase inhibitors in patients as they reach the moderate to severe stage of AD but there are no clinical trials under way that will provide this evidence. There is also a need to evaluate the use of memantine, which already has a licence for the treatment of moderate to severe AD, alone and in combination with a cholinesterase inhibitor at the point where patients make the transition to moderate to severe disease. To most effectively inform clinical practice and relevant decision-making bodies such as NICE, any trials that examine these issues should be pragmatic and based on a representative patient population in whom clinicians currently have genuine uncertainty about how to proceed with treatment. Important outcome measures for such a trial should include preservation of activities of daily living and independence as well as cost-utility data. By recruiting patients who have reached the transition point between mild-tomoderate and moderate-to-severe dementia (SMMSE 10-12) and who are already receiving anticholinesterase treatment, DOMINO is designed to Page 17 of 68

answer clinical questions about efficacy and cost-effectiveness that have real relevance to clinicians and policy making organisations such as NICE. 4.2 Clinical Studies A number of studies have demonstrated that acetylcholinesterase inhibitors modestly improve cognition in a subgroup of patients with mild to moderate AD and stabilize cognition, function and behaviour for 6 months (Rogers 1998a, 1998b; Corey-Bloom 1998; Burns 1999; Rosler 1999; Raskind 2000; Tariot 2000; Wilcock 2000; Farlow 2000; Rockwood 2001) and may continue to exert benefit for 12 months (Mohs 2001; Winblad 2001). Evidence has also been presented to suggest that these drugs may continue to have a beneficial effect for up to 2 years (AD2000 Collaborative Group 2004) although a recent influential systematic review concluded that benefits of treatment are modest and that the methodological quality of most published trials could be questioned (Kaduszkiewicz 2005). Although there are no adequate systematic reviews of the use of acetylcholinesterase inhibitors in patients with moderate to severe AD, acetylcholine and choline acetyltransferase levels do not begin to fall significantly until dementia is advanced (Davis 1999; Gilmor 1999). Hence, there would be good theoretical reasons to anticipate a therapeutic effect of cholinesterase inhibition at later stages of disease than the currently licensed indication of mild to moderate severity. Trial evidence for this comes from four main sources: 1) Randomised double-blind placebo-controlled trials that have included moderate to severely affected patients. Feldman (2001) showed significant benefits with donepezil treatment over 24 weeks in cognitive, behavioural and functional outcomes in a group of patients whose SMMSE scores ranged from 5-17 (mean score for patients receiving donepezil = 11.7). 2) Secondary analyses of trial data from mild to moderate patients examining the subgroup of patients with more severe illness. Wilkinson (2002) performed a post hoc analysis on pooled data from 124 patients with SMMSE scores of 10-12 from the 4 pivotal galantamine studies. Cognitive and functional abilities were significantly improved in galantamine treated participants. Burns (2004) retrospectively analyzed pooled data from 117 patients selected from three RCTs of rivastigmine on the basis of a SMMSE score of 10-12 points. Rivastigmine treatment over 6 months showed significant benefits in cognitive and behavioural domains. Finally, Gauthier (2003) and Feldman (2005) reanalyzed the Feldman (2001) data focussing on 145 patients with a SMMSE score of 5-12. At week 24 LOCF the mean differences in mean change from baseline scores were 2.0 for the SMMSE and 7.4 for the SIB in favour of the donepezil treated participants and CIBIC-plus scores were significantly improved compared with placebo with a 0.70 mean treatment difference. 3) Long term trial data within which participants have progressed to more severe disease stages of illness can also demonstrate apparent continued efficacy in moderate to severe patients. In the AD2000 trial, (AD 2000 collaborative group 2004) 49% of participants randomised to treatment with Page 18 of 68

donepezil had SMMSE scores of 10-18 points at study entry. Over the 2-year study period, the donepezil group averaged SMMSE scores 0.8 points higher than the placebo group. This benefit was not restricted to any subgroup of patients in terms of severity rating and was maintained over the trial period. Raskind (2004) demonstrated benefits of continuing galantamine treatment for 36 months in a group of patients whose mean SMMSE score at entry had been 19.7 points and a proportion of whom would have entered the moderate to severe category during the course of the study. 4) Trials examining the effects of cholinesterase inhibitor withdrawal. Typically, following a placebo washout period of 6 weeks at the end of a trial, the benefits of 24 weeks of donepezil treatment in terms of cognition and global function are lost (Doody 2001). Most of such washouts have been at the end of relatively short trials at which point patients have still been only mildly to moderately affected. There are no published randomised control trials examining the effects of treatment withdrawal in patients at the mild to moderate/moderate to severe boundary that would help clinicians to make decisions at the point where NICE guidance advises stopping. Holmes (2004) showed behavioural benefits of continuing donepezil in a group of patients selected on the basis of marked neuropsychiatric symptoms (NPI score>11) at baseline and with a mean SMMSE score of 21.1 points. There is intriguing preliminary evidence that a combination of a cholinesterase inhibitor and memantine might be particularly beneficial in patients at this severity point. For example, a study of 404 patients with moderate to severe AD (SMMSE 5-14) who had been stabilised on donepezil treatment for at least 6 months, investigated the effect of adding memantine 10mg b.d. for 6 months (Tariot 2004) The change in total mean (standard error) scores favoured memantine over placebo for the SIB; 1.0 (0.7) vs. 2.4 (0.74), ACDS-ADL19; -1.7 (0.51) vs. 3.3 (0.55) and CIBIC-Plus; 4.38 (0.081) vs. 4.64 (0.087). All other secondary measures showed significant benefits of memantine treatment and treatment discontinuations because of adverse events were seen in 7.4% receiving memantine and 12.4% in those receiving placebo. An unpublished Phase III study in mild to moderate AD patients, however, has suggested that combinations of a variety of cholinesterase inhibitors and memantine may not provide additional benefits over monotherapy in this less impaired group (Scrip 2003). 4.3 National Institute for Clinical Excellence Guidance The National Institute for Clinical Excellence (NICE) recommended in 2001 that cholinesterase inhibitors (donepezil, rivastigmine and galantamine) should be offered to patients with mild to moderate AD whose SMMSE score was above 12 points. NICE guidance from 2001-2005 has been that prescription should only be continued while the SMMSE score remains above 12 points and the patient s global, functional and behavioural condition remains at a level where the drug is considered to have a worthwhile effect. This recommendation was made on the grounds of cost-containment, rather than clinical efficacy, since many of the patients who entered the trials that established efficacy had SMMSE scores of as low as 10 points. Page 19 of 68

The current NICE Guidance recommends that the three acetylcholinesterase inhibitors donepezil, galantamine and rivastigmine are used as options in the management of people with AD of moderate severity (SMMSE score of between 10 and 20 points). Patients continuing on the drugs should be reviewed six monthly by SMMSE score and global, functional and behavioural assessment, incorporating the views of carers. Drug treatment should only be continued while the patient s SMMSE score remains above 10 points and their global, functional and behavioural condition remains at a level where the drug is considered to be having a worthwhile effect. Memantine is not recommended as a treatment option for people with moderately severe to severe Alzheimer s disease except as part of a clinical study. 4.4 Choice of Study Population This study has been designed to be pragmatic with broad entry criteria which should facilitate recruitment of a representative clinical population, thereby maximising the generalisability of the data. Diagnostic criteria will include both probable and possible AD (McKhann et al 1984), with exclusion criteria kept to the minimum required to maintain safety and to avoid confounding factors. 4.5 Choice of Investigational Drugs Donepezil is the most frequently prescribed cholinesterase inhibitor in the UK, has once daily dosage and a favourable side effect profile (see section 10.6 for full side effect profile). Memantine is the only drug licensed for the treatment of moderate to severe AD; it is well tolerated and may have a synergistic effect with donepezil (Tariot 2004). 4.6 Choice of Outcome Measures Dementia due to Alzheimer s disease is a complex, multi-faceted disorder. However, more generalized cognitive impairment leading to functional disability are core components. Consequently DOMINO-AD will assess change in both domains as primary outcome variables. Though several instruments are available to measure such change, cognitive function will be measured with the SMMSE and activities of daily living will be measured with the BADLS. Both are well validated instruments and although the SMMSE is less sensitive to change than other more detailed measures such as the Severe Impairment Battery within the dementia severity range under study, the SMMSE has been used in so many studies that its inclusion is important to allow comparisons with earlier trial data and to increase generalisability of DOMINO s findings. Secondary outcome measures will include NPI (measuring non-cognitive dementia symptoms), EQ-5D, DEMQOL-proxy (both measuring health related quality of life), GHQ-12 (measuring care-giver burden), cost effectiveness (the combination of CSRI with BADLS, DEMQOL and EQ-5D) and institutionalisation. See section 9 for further information The interview schedule is based on other successful trials in dementia (e.g. CALM-AD) and is designed to last no longer than 1.5 hours for the patient and for the carer. Page 20 of 68

4.7 Consumer Involvement This study has been designed in collaboration with the Alzheimer s Society. As part of the decision to co-fund the DOMINO study, the protocol was reviewed by the Alzheimer Society s Quality Research in Dementia (QRD) consumer network. This is a group of 150 carers, former carers or people with dementia who systematically review all research applications to the Alzheimer s Society, and were very positive about the importance of this study. Two members of the QRD network have assisted with the development of the study information sheets. A QRD member will also sit on the trial steering group. 4.8 Risks and Benefits 4.8.1 Potential Risks As both donepezil and memantine are well tolerated drugs, the main risk for participants is withdrawal of their existing prescription of donepezil. Patients will only be entered into the study if, following assessment and discussion with the patient and carer and taking into account the NICE guidance, there is substantial uncertainty that they are benefiting from ongoing donepezil treatment. Therefore they would be likely to be taken off donepezil if they were not on the trial. Our use of a graded withdrawal of donepezil, with a 50% dose reduction for 4 weeks followed by complete withdrawal, and provision of clinical supervision and assessment of participants during this period should minimize potential harm. It should be borne in mind that all of the participants would be considered ineligible for cholinesterase inhibitor treatment under even the 2001 NICE guidance, so that withdrawal of treatment cannot be considered unethical or against accepted best practice in England and Wales. Similarly, memantine will be prescribed on an escalation dose basis for the first 4 weeks to ensure any adverse reactions are minimised. The research assessments will take some time, but the patients will have the option to have these in their home to minimise inconvenience and distress. 4.8.2 Potential Benefits Though neither donepezil nor memantine are recommended by NICE for the treatment of patients with low SMMSE scores on the basis of cost effectiveness, both drugs have shown advantages over placebo on important clinical outcomes in this patient group (see section 4.2 clinical studies for a more detailed explanation). Similarly the drugs have been shown to be well tolerated. Consequently participants randomised to receive active drug(s) may potentially benefit from an improvement or a reduction in deterioration of their symptoms. Page 21 of 68

5 TRIAL OBJECTIVE AND PURPOSE 5.1 Aim The aim of the DOMINO study is to determine, in a factorial (2x2) design whether there is worthwhile benefit for patients for whom there is uncertainty on whether or not to continue cholinesterase inhibitors from: 1) adding memantine to donepezil, 2) switching to memantine, or 3) continuing donepezil or 4) placebo. Memantine Donepezil Continue Add Arm 1 Donepezil Memantine No Arm 3 Donepezil Memantine placebo Table 2: description of each arm of the study Discontinue Arm 2 Donepezil placebo Memantine Arm 4 Donepezil placebo Memantine placebo 5.2 Objectives 5.2.1 Primary Objectives The trial will test a number of hypotheses in memory clinic patients who have declined in terms of cognitive function to reach the transition point to moderate-to-severe AD. a) Patients with AD who continue donepezil (Groups 1 and 3) beyond the moderate to severe transition point will show a significantly smaller decline on ratings of cognitive function and activities of daily living over the following 12 months than those discontinuing donepezil (Groups 2 and 4). Analysis at the margins using 100% of trial participants. b) Patients with AD who commence memantine therapy (Groups 1 and 2) at the moderate to severe transition point will show a significantly smaller decline on ratings of cognitive function and activities of daily living over the following 12 months than those who do not (Groups 3 and 4). Analysis at the margins. c) Patients given the combination of memantine and donepezil (Group 1) at the moderate to severe transition point will show additive or synergistic significant benefits on measures of activities of daily living and cognitive function after 12 months compared to those patients continuing on either monotherapy (Groups 2 and 3). Test for interaction within table. 5.2.2 Secondary Objectives a) Patients with AD who continue donepezil (Groups 1 and 3) beyond the moderate to severe transition point will show a significantly smaller deterioration on ratings of non-cognitive symptoms and health related quality of life over the following 12 months than those discontinuing Page 22 of 68

donepezil (Groups 2 and 4). Analysis at the margins using 100% of trial participants. b) Patients with AD who commence memantine therapy (Groups 1 and 2) at the moderate to severe transition point will show a significantly smaller deterioration on ratings of non-cognitive symptoms and health related quality of life over the following 12 months than those who do not (Groups 3 and 4). Analysis at the margins. c) Patients given the combination of memantine and donepezil (Group 1) at the moderate to severe transition point will show additive or synergistic significant benefits on measures of non-cognitive symptoms and health related quality of life after 12 months compared to those patients continuing on either monotherapy (Groups 2 and 3). Test for interaction within table. d) Treatment of patients with donepezil beyond the moderate to severe transition point will be more cost-effective than discontinuing donepezil. Memantine therapy will be more cost-effective than placebo. The combination of memantine and donepezil will be more cost-effective than monotherapy. e) Patients who continue on donepezil beyond the moderate to severe transition point will be institutionalised later than those who do not. Patients who commence memantine therapy will be institutionalised later than those taking placebo. Patients who commence combination of memantine and donepezil will be institutionalised later than those on monotherapy. For carers, parallel secondary objectives will be changes in psychological morbidity and health related quality life. Page 23 of 68

6 TRIAL DESIGN 6.1 Design This will be a pragmatic, multi-centre, double-blind, randomised, placebocontrolled (double dummy), parallel group, 2X2 factorial clinical trial. 6.2 Trial Flow Chart S C R E E N I N G POPULATION Patients under care of 15 older adult services with a diagnosis of Alzheimer s disease ELIGIBILITY Donepezil 3/12 on 10mg 6/52, SMMSE 5-13, uncertaintity re continued benefit of donepezil, community dwelling, carer CONSENT Patient or Personal Legal Representative BASELINE ASSESSMENT EXCLUSIONS On memantine Contraindication to trial drugs or concerns over compliance Severe or unstable illness Patient already in a trial RANDOMISATION MRC CTU Donepezil 10mg Memantine 20mg Donepezil Placebo Memantine 20mg Donepezil 10mg Memantine Placebo Donepezil Placebo Memantine Placebo 6 week assessment 18 week assessment 30 week assessment 52 week assessment Follow up every 6/12 for 3 years Page 24 of 68

6.3 Trial Duration 6.3.1 Duration of the treatment period The proposed duration of the treatment period is 52 weeks post randomisation. 6.3.2 Duration of the follow-up period Study measures will be assessed prior to randomisation, at week 6 to assess the acute effects of donepezil withdrawal, at week 18, week 30 and at week 52. Participants will then be followed up every 26 weeks for 3 years by telephone interview to establish whether and on what date they entered a care institution. 6.3.3 Definition of completion of the trial for an individual participant Completion is defined as completion of 52 weeks on the trial medication or discontinuation of follow-up for any cause. Participants who discontinue taking the trial medication should remain in follow-up if they are in agreement to do so. Participants may not formally discontinue their follow-up and remain on the trial medication. 6.3.4 Definition of the end of the trial In regulatory terms, the end of the trial is defined as the end of the interventional phase, 52 weeks after the last patient is randomised. For the purposes of the ethics authorisations, the end of the trial is defined as 4 years after the last patient is randomised to allow for the 3 year long phase of long term follow-up and outcome data collection. 6.4 Premature Trial Closure The trial may be stopped by the Trial Steering Committee. The Independent Data Monitoring Committee in accordance with the IDMC charter may recommend to the Trial Steering Committee that the trial be stopped. The criteria for stopping the trial will be established as part of standard operating procedures of the IDMC prior to the commencement of the trial and patient recruitment. Page 25 of 68

7 SELECTION AND WITHDRAWAL OF SUBJECTS 7.1 Number and Source of Participants The trial aims to recruit 800 participants who will be drawn from referrals to, and other contacts with, old age psychiatric services in England; these will include community mental health teams and their associated memory clinics. The following 15 clinical recruiting centres will be involved: Bath: Avon and Wiltshire Mental Health Partnership NHS Trust (Professor Jones) Belfast: Belfast Health and Social Care Trust (Dr Passmore) Birmingham: Birmingham and Solihull Mental Health NHS Trust (Dr Bentham) Cambridge: Cambridgeshire and Peterborough Mental Health Partnership NHS Trust (Dr Dening) Dundee: Dundee Community Health Partnership (Dr Findlay) Glasgow: Greater Glasgow and Clyde Health Board (Dr Hughes) Leicester: Leicestershire Partnership NHS Trust (Professor Lindesay) London Maudsley: South London and Maudsley Foundation NHS Trust, London (Professors Howard, Ballard, Banerjee) London Imperial: West London Mental Health NHS Trust (Dr Ritchie) Manchester: Manchester Mental Health and Social Care NHS Trust (Professor Burns) Newcastle: Northumberland Tyne and Wear NHS Trust (Professors McKeith, O Brien) Nottingham: Nottinghamshire Healthcare NHS Trust (Professor Jones) Oxford: Oxfordshire and Buckinghamshire Mental Healthcare Partnership NHS Trust (Dr McShane and Professors Jacoby and Wilcock) Southampton: Warwick: Hampshire Partnership NHS Trust (Professor Holmes) Coventry and Warwickshire Partnership NHS Trust (Dr Sheehan) 7.2 Planned recruitment rate Each centre will be expected to recruit 55 patients over 2 years. This is 2-3 per centre per month which, from review of current eligible patients within the memory and old age clinics of the above centres is achievable. Month Target per Overall target Target per Overall target centre recruitment centre recruitment (cumulative) (cumulative) 2 2 30 2 30 4 3 45 5 75 6 5 75 10 150 8 5 75 15 225 10 5 75 20 300 12 5 75 25 375 14 5 75 30 450 16 5 75 35 525 18 5 75 40 600 20 5 75 45 675 Page 26 of 68

22 5 75 50 750 24 5 75 55 825 Total 55 825 55 825 Table 3: Recruitment Targets 7.3 Recruitment Strategy A research worker will be employed at each site who will recruit and assess patients as well as drawing on help from DeNDRoN and MHRA who have adopted the trial and are willing to provide help with publicising the trial, training and recruitment. Participants will be identified from patients with Alzheimer s disease meeting study entry criteria who are being followed up in memory clinics, out patient clinics or other components of specialist mental health, geriatric medicine or neurology services. Once a potential participant has been identified, the clinician will obtain their verbal consent to be pass their details to the research worker. The research worker will then recruit the patient in line with the trial standard operating procedures. In some centres it may be necessary to recruit patients referred to the trial from the community rather than those being followed up in memory clinics, out patient clinics or other components of specialist mental health, geriatric medicine or neurology services. This will most likely be in areas where, due to local health service policies, patients are initiated on donepezil treatment by a consultant but then referred back to their GP for all follow-up care. In these cases, patients will be referred to the PI or RW of their local centre by their GP if, in the opinion of the GP, the patient has reached the point at which there is uncertainty that they are receiving benefit from a continued prescription of donepezil and the patient verbally consents to the referral. Whilst the GPs would be encouraged to also check other eligibility criteria prior to referral, it will be the responsibility of the PI (or delegated staff as appropriate) at the recruiting centre to gain consent, establish eligibility and carry out all trial related procedures. 7.4 Monitoring and ensuring recruitment to the trial Recruitment will be monitored by the TMG, the TSC and the IDMC. The intention will be to identify problems early and problem-solve to bring recruitment back on track. The local recruitment frames are the same size and there will be careful monitoring and support to maximise recruitment. We believe that these factors will minimise the likelihood of failure to recruit in individual centres and overall. We have built in a stepped increase into the recruitment targets during the first 4 months to allow for any teething problems and learning curves. 7.5 Consent Procedure Where possible, fully informed consent will be obtained from the patient. However patients entering this study will have moderate to severe dementia with SMMSE scores of between five and thirteen and the majority of patients with this degree of dementia will lack the necessary mental capacity to give fully informed consent. The aims of the study are incompatible with entering Page 27 of 68