INSTITUTE FOR CLINICAL AND ECONOMIC REVIEW Lecanemab for Early Alzheimer's Disease Final Evidence Report April 17, 2023 Prepared for =CTAF EEEEDT] CALIFORNIA TECHNOLOGY ASSESSMENT FORUM ©Institute for Clinical and Economic Review, 2023 AUTHORS: Grace Lin, MD Medical Director for Health Technology Assessment Institute for Clinical and Economic Review Melanie D. Whittington, PhD, MS Director of Health Economics Institute for Clinical and Economic Review Abigail Wright, PhD, MSc Senior Research Lead, Evidence Synthesis Institute for Clinical and Economic Review Foluso Agboola, MBBS, MPH Vice President of Research Institute for Clinical and Economic Review Serina Herron-Smith, BA Associate Research Manager Institute for Clinical and Economic Review Steven D. Pearson, MD, MSc President Institute for Clinical and Economic Review David Rind, MD, MSc Chief Medical Officer Institute for Clinical and Economic Review DATE OF PUBLICATION: April 17, 2023 OInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Pagei How to cite this document: Lin GA, Whittington MD, Wright A, Agboola F, Herron-Smith S, Pearson SD, Rind DM. Beta-Amyloid Antibodies for Early Alzheimer's Disease: Effectiveness and Value; Evidence Report. Institute for Clinical and Economic Review, April 17, 2023. https://icer.org/assessment/alzheimers-disease-2022 /#timeline Grace Lin served as the lead author for the report. Abigail Wright led the systematic review in collaboration with Serina Herron-Smith. Foluso Agboola provided methodologic guidance. Melanie Whittington developed the economic model. David M. Rind and Steven D. Pearson provided methodologic guidance on the clinical and economic evaluations. We would also like to thank Noemi Fluetsch, Monica Frederick, Kelsey Gosselin, and Yasmine Kayali for their contributions to this report. Additionally, we would like to thank UsAgainstAlzheimer's for connecting us with individuals living with Alzheimer's disease and caregivers to help us better understand their lived experiences. ©lnstitute for Clinical and Economic Review, 2023 Page ii Alzheimer's Disease - Final Evidence Report About ICER The Institute for Clinical and Economic Review {ICER} is an independent non-profit research organization that evaluates medical evidence and convenes public deliberative bodies to help stakeholders interpret and apply evidence to improve patient outcomes and control costs. Through all its work, ICER seeks to help create a future in which collaborative efforts to move evidence into action provide the foundation for a more effective, efficient, and just health care system. More information about ICER is available at https://icer.org/. The funding for this report comes from government grants and non-profit foundations, with the largest single funder being the Arnold Ventures. No funding for this work comes from health insurers, pharmacy benefit managers, or life science companies. ICER receives approximately 25% of its overall revenue from these health industry organizations to run a separate Policy Summit program, with funding approximately equally split between insurers/PBMs and life science companies. No life science companies relevant to this review currently participate in this program. For a complete list of funders and for more information on ICER's support, please visit https://icer.org/who-we-are/independent-funding/. For drug topics, in addition to receiving recommendations from the public, ICER scans publicly available information and also benefits from a collaboration with IPD Analytics, an independent organization that performs analyses of the emerging drug pipeline for a diverse group of industry stakeholders, including payers, pharmaceutical manufacturers, providers, and wholesalers. IPD provides a tailored report on the drug pipeline on a courtesy basis to ICER but does not prioritize topics for specific ICER assessments. About CTAF The California Technology Assessment Forum (CTAF) - a core program of ICER - provides a public venue in which the evidence on the effectiveness and value of health care services can be discussed with the input of all stakeholders. CTAF seeks to help patients, clinicians, insurers, and policymakers interpret and use evidence to improve the quality and value of health care. The CTAF Panel is an independent committee of medical evidence experts from across California, with a mix of practicing clinicians, methodologists, and leaders in patient engagement and advocacy. All Panel members meet strict conflict of interest guidelines and are convened to discuss the evidence summarized in ICER reports and vote on the comparative clinical effectiveness and value of medical interventions. More information about CTAF is available at https://icer.org/who-we-are/people/independent-appraisal-committees/ctaf. The findings contained within this report are current as of the date of publication. Readers should be aware that new evidence may emerge following the publication of this report that could potentially influence the results. ICER may revisit its analyses in a formal update to this report in the future. The economic models used in ICER reports are intended to compare the clinical outcomes, expected costs, and cost-effectiveness of different care pathways for broad groups of patients. Model results therefore represent average findings across patients and should not be presumed to represent the clinical or cost outcomes for any specific patient. In addition, data inputs to ICER models often come from clinical trials; patients in these trials may differ in real-world practice settings. ©Institute for Clinical and Economic Review, 2023 Page iii Alzheimer's Disease - Final Evidence Report In the development of this report, ICER's researchers consulted with several clinical experts, patients, manufacturers, and other stakeholders. The following experts provided input that helped guide the ICER team as we shaped our scope and report. It is possible that expert reviewers may not have had the opportunity to review all portions of this report. None of these individuals is responsible for the final contents of this report, nor should it be assumed that they support any part of it. The report should be viewed as attributable solely to the ICER team and its affiliated researchers. For a complete list of stakeholders from whom we requested input, please visit: https://icer.org/assessment/alzheimers-disease-2022/#overview Expert Reviewers VICTOR HENDERSON, MD, MS Professor Stanford University No relevant conflicts of interest to disclose, defined as more than $10,000 in healthcare company stock or more than 55,000 in honoraria or consultancies during the previous year from health care manufacturers or insurers. Dr. Henderson has previously received honorarium for serving as an external reviewer on an ICER review in 2021. JASON KARLAWISH, MD Professor University of Pennsylvania Dr. Karlawish has served as a site investigator for clinical trials sponsored by Biogen, Eisai Co., Ltd and Eli Lilly & Co. We also received comments from consultants of a patient organization but have not named them at their request. Those comments did result in some changes to the Report. ©lnstitute for Clinical and Economic Review, 2023 Page iv Alzheimer's Disease - Final Evidence Report Table of Contents L Lol AV Uy o T 1 R = 7 Yol 4 o T T 1 2. Patient and Caregiver PerspeCtiVes ...... ... ciieireeenieeetercseeeseeese st e eeee e e s ee e st e sees e s neaesameeeenneens 4 3. Comparative Clinical EffeCtiVENESS ......cuvvirvrerrircieieeeveeerrrrrceerrrscsersrssssseeesseessnressssssessssssassessssnneneses 8 3.1. MethodSs OVEIVIEW........ccciiiiiieititrte ettt 8 SCOPE OF REVIEW ....ueeiiiiiciiiierisieirerrse e e s rrcnseeeseineeeresssssssesssssseeesssaneeesessnsesessasssnesrsssaneessnsnneereesnsens 8 EVIENCE BASE ..ottt e s 8 0 0T U ] 10 CliNICal BENETILS ... .o 10 = 10 T 12 Subgroup Analyses and Heterogeneity.......cccucceuccirinriircce e rcr e e s e e s e 14 Evaluation of Clinical Trial DIVersity......cccccuerteeceriininieeseesse st e sseesseeeesees e se e s seesene s e s seesnas 16 Uncertainty and CONTIOVEISIES .....uevvveveierirciereircisirnrerissssrerisssssrereesssseessesssnessssssssesssssaseesessssessen 16 3.3. SUMMAry and COMMENT ...t rceere e e e s e s e e st e s e e s e en e s e e se e e s e mn e s eneaenmnennanens 19 O 1 Y o = 21 4. LoNg-Term CoSt EffeCtiVENESS. .....cccceiciieiieeciercieeieecie st cste s seesese et esseessse e st e s eessse s e e e s seessnneennanneans 22 4.1. Methods OVEIVIEW.........coimiimiiiiiiectre ettt s s s s s sr e e 22 4.2. Key Model Assumptions and INPULES ......c.coiciciiniiriereen et rcrercsee e ee e s e s e 24 I TR 0] ] 3 R 25 BasS@-Case RESUILS.......ccciuiiiiiiiii ittt s et s b s s s s s n s s 25 SENSILIVIEY ANAIYSES .ceiiiveeirie et e irc et rr et s e rrcsste e s ss e e e e ss et e e ssssse e e e ss s e e eesesssneeessssaeeeessssnneesnanns 26 SCENAMO ANAIYSES....eeieieciirierie e e rrerer e e re v e e rrcraee e s sssssn e e e ssrseeeerss s e e e eaesnaseeasssssneesssssnneeesssnneeseanas 28 ThresShold ANAIYSES .........coiiieiece ettt r e e st e s rer e s ses e et e e ee st e e s an e s eneaeneneasaaneennas 29 Model ValidatioN.........cociviiiiicere et 30 Uncertainty and CONTIOVEISIES .....cuvvvevreerrirrrereircisrnresrsrnerrisesssresssssseeessessssnsssssssssesssssasesssssaseesss 30 4.4 SUMMArY aNd COMMIEBNT ....ueiierieeereirereererrcreeeeresssseerssssnresssssserreasssssessessssnsesssssnresssssasesessssssesss 31 5. Contextual Considerations and Potential Other Benefits.........ccocvnincnininnnccccicns 32 O 17 YL 34 6. Health Benefit Price BeNChMArks .........cooeiiiiiiiiniiii et 36 O©Institute for Clinical and Economic Review, 2023 ~ Pagev Alzheimer's Disease - Final Evidence Report L Y Y o ) N 36 7. Potential BUAEEt IMPACT .....cooi et ese e s e s e s en e s e e e s 37 7.1. Overview Of K&y ASSUMPLIONS .....cccceeeveeeiercrieniirrsiesssessseesseesssesessesseesasessssnssssesssssansesssassssssnsenas 37 720 U= PRI 37 Access and Affordability AlErt.........oo et s ee s 39 8. Policy RECOMMENAAtIONS. ...coiiiiie ettt eeccer s e s e e e s rcr e s e e e s e e e s smn e s en e s e e s e me e s nens 40 All SEAKENOIENS ...t 40 MuUItiple StaKEOIAETS ... .o ettt e et e et e e e st e e st e s e e e et e e e ame e e aes 44 P OIS .. ettt R e e aaE R aa R R R e R e s e e n e aaan 46 MANUFACTUTETS ...ttt e e s ae e s n e sen e e s ene e s ne e 46 REFEIENCES ...ttt sttt st st s e s e e e s e e et e et s se st e e as e b e e asaneenssnenne s 49 A.Background: Supplemental INfOrmation .......cceviecvrieiriierrrrrrretrrrrcrree e rcreeee s rreerrrssrsnee s s sessesesenanns Al AL DEfiNItIONS....coeeiieeeccit et Al A2. Potential Cost-Saving Measures in Alzheimer's DiSEase..........ccevereeerseerseereeeesieeceeensseneseesseens A6 B.Patient Perspectives: Supplemental INformation...........coeoeeoeeonerecee e B1 23 1V 1= ] o T L Bl C.ClINTCAl GUIAEINES ...t Cc1 American Academy Of NEUIOIOBY ......coeimieiieiriiececceeeeeeie s s s s se e s e s e s e sse e e snssesssnsnns C1 National Institute for Health and Care Excellence (NICE)® .........c.ooeeeerrmrmrmrmrmrcrecereeeeeeeens C1 American Psychiatric ASSOCIAtION..........coveieceieicieceee et se s se e e e e e seeseen Cc1 The National Institute on Aging-Alzheimer's Association"7492 . ........cooeeevvrercerimrsesinsnene c2 Aducanumab: Appropriate Use Recommendations® ...........c.ccoeoerecriececcncccecceec e C2 D.Comparative Clinical Effectiveness: Supplemental Information.......c.ccccccceeiimrninciencenceecseesceenee D1 D1. Detailed Methods........ccccoiiiiiieeiiietctcccr et sse s sn s s D1 PICOTS ...ttt s s s et s s n e s s s e s s e e et s e et e s e s nnsansnnnns D1 Data Sources and SEarches ..o s D6 STUAY SEIECHION. ...ttt ettt s e e et r st e s es e st e e e san e s eaenne e e snmeenananenans D9 Data EXEraCioN. .. ..ciiiiiiiiiniiiiicn ettt D10 Risk Of Bias ASSESSIMENT ...c...eviieiiceciiiriiit ettt e e s e s sms s D10 Evaluation of Clinical Trial DIVErSity.......cccccereereceeeieercieseeeceeresresseesseeseeesseessseseseesseesssnessseassnans D12 OInstitute for Clinical and Economic Review, 2023 Pagevi Alzheimer's Disease - Final Evidence Report RESUIES c.ceveeeeeeeeeressesssssssseseee eeeeeeeeesssssssssssreseseesesaassssssssssssssssssssssteeeeeeteeeeetsessesseessserreeseeeeeeeeeeeenees D13 D2. Additional Clinical EVIdeNCE........ooeiiciicicicicicnii st D15 [I0=T or= g T=T 0 - T « D15 D3. EVIdENCE TaDIES ... D27 DA. ONEOINEG STUAIES ... cciceicrirrrirrierre e sceretessteseer st eeseesssasese s s e e s sns s s se s aasne e e st esssesansessnsassnessnns D66 D5. Previous Systematic Reviews and Technology Assessments .......cccceeeciereriercsercccmrrninnsncnen. D68 1 D68 Pang, M. et al. (2022). "Effect of Reduction in Brain Amyloid Levels on Change in Cognitive and Functional Decline in Randomized Clinical Trials: An Instrumental Variable Meta-Analysis">°D68 Lacorte, E. et al. (2022). "Safety and Efficacy of Monoclonal Antibodies for Alzheimer's Disease: A Systematic Review and Meta-Analysis of Published and Unpublished Clinical Trials"#........ D69 Villain, N. et al. (2022). "High-clearance Anti-amyloid Immunotherapies in Alzheimer's Disease. Part 1: Meta-analysis and Review of Efficacy and Safety Data, and Medico-economical ASPECES"" 1120 ettt s b e s b e b e b e b e b e b e b e e s ea s e b s eas et b e b et e benbenes D70 E.Long-Term Cost-Effectiveness: Supplemental INnformation........ccceeeeevvrciereenecsinenieseeevreceeeerecneens El E1l. Detailed Methods........cccooiiii et s El BT {10 2 o 10 - 14 o o E2 TreatMENt StratEgIeS . oo e e s s s e s s e s s s e s e e s s s e s e s s e e e s e e e e e e e e e e e e e e e e e e et e e e e e e e e e nnnnnnn s E2 E2. Model Inputs and ASSUMPLIONS .....cciiireicrerieimrrersciereercieeeerssssssresissssrsrrsesssseesessssnesssssssssssesssenes E3 MOEI INPULES ...ttt rcr e e e s s s e st e e e s ee e s s e e s e me e e sen e s e e e ameenenanens E4 235 TR =TT U R E13 Description of @VLY CalCUIQtioNs ........cocveeceeriiieiieeerercte et seeesee e s e eessse s se e s e e s sanesaeeas E13 E4. SENSItIVItY ANAIYSES ....ooeeeieeeeccrcciie e e s s ee e e s s s sen e s e e e e e s e ne e e e neenens E13 E5. SCENATIO ANGIYSES.......eiiicciereieecirercaire s e e s eeese et e s e es et e eeneesesanesenceensaseasasme s s seneneneasannsanenerens E14 Scenario Analysis 1: Treatment Stop at Severe AD....... ..o iircieieeieeecre et creeee e e e s er e E14 Scenario Analysis 2: Updated Caregiver Disutility EStimates........cccccecceerrmrnrrrrcersrnrsnsnenssensns E14 E6. MOdel Validation.......c.coveiiieeieecetiee e E15 Prior ECONOMIC MOMEIS.......coiiiiiiiiiiisi i E15 F.Supplemental Policy RecommENndations.........ccooiieieiiirieciee et e e ee e e e e e e s F1 PAYEIS ... e e s s e e e e e e e e e e et e e e e e e e e e e e e e e e e e e eeeeeeeeee e ettt ee-eeeeeeeeeeteeeeeeeeteyaaaaaaaaaaaaaaaan F1 Coverage Criteria: GENEIAl......iivveerrieciiiriireeceeerrsreerrrssrereessissssresrssssessrsssssresssssneeesessssnnerssssansersen F1 OInstitute for Clinical and Economic Review, 2023 Pagevii Alzheimer's Disease - Final Evidence Report Drug-Specific Coverage Criteria: LeCanemab........ccccceereeermierircieieircireeeennnrenesissssseesvssssessssessesesss F1 G.PUDIIC COMMENTS.....ccciiiiiiiiiiiiieeeeeeeeececccciierere e ses s s assssaaeaaaanaanansssssssssssssassssssssssssssssssnnssssssssssssssnnnanns Gl H.CONTlICt Of INTEIESt DISCIOSUIES....ccciiiiiiiiccceiieeeee e cecce s e e e s e e s e s s ssss s e s eeessassasssssnnssesasssassssssnnsseeeessnss H1 ©lnstitute for Clinical and Economic Review, 2023 Page viii Alzheimer's Disease - Final Evidence Report List of Acronyms and Abbreviations Used in this Report AD ADAS-Cog 13 ADSC-ADL AE AHRQ APOE €4 ARIA[-E,H] CDR-SB cl CSF evLYG FDA HBPB HR HRQolL ICER MMSE MRI N n NICE NPI-10 NR PET PICOTS PV4 QALY SD SE SUVR us WAC Alzheimer's disease Alzheimer's Disease Assessment Scale - Cognitive Subscale Alzheimer's Disease Cooperative Study - Activities of Daily Living Inventory Adverse event Agency for Healthcare Research and Quality Apolipoprotein €4 Amyloid-related imaging abnormalities[-edema/effusion, hemorrhage or superficial siderosis] Clinical Dementia Rating Scale - Sum of Boxes Confidence interval Cerebrospinal fluid Equal value of life years gained Food and Drug Administration Health benefit price benchmark Hazard ratio Health-related quality of life Institute for Clinical and Economic Review Intention-to-treat Intravenous kg Kilogram Long-term care Life year Mild cognitive impairment Milligram Mini-Mental State Exam Magnetic resonance imaging Total number Number National Institute for Health and Care Excellence Neuropsychiatric Inventory 10 Not reported Positron emission tomography Population, Intervention, Comparators, Outcomes, Timing, Settings Protocol Version 4 Quality-adjusted life year Standard deviation Standard error Standardized uptake value ratio United States Wholesale acquisition cost OInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page ix Executive Summary Alzheimer's disease (AD) is a fatal neurodegenerative brain disease characterized by the progressive accumulation of beta-amyloid protein plaques and neurofibrillary tangles; these are hypothesized to damage neurons and lead to the loss of cognition and physical functioning.! AD affects more than six and a half million people in the United States (US), with more women than men affected and Black Americans at a higher risk of developing the disease.? Symptoms of AD include impairment of memory, language, executive function, and visuospatial function that affects one's ability to care for themselves. People living with AD require a substantial amount of caregiving, and eventually may require around-the-clock in-home or institutional care. Caregivers, most often unpaid family members and friends, can suffer significant negative physical, financial, and emotional outcomes from the strain of caregiving.3* Current treatment of AD is focused on supportive care, including treatment of dementia symptoms with medications that do not alter the course of the disease.>® Because of the devastating burden of AD, there is a great need for disease-modifying treatments (DMTs) that slow or stop progression of the disease. Although aducanumab (Aduhelm™, Biogen) was granted accelerated approved for the treatment of AD in June 2021, there remain substantial uncertainties about its benefits and harms. In this report, we focus on lecanemab (Legembi™, Eisai Co., Ltd), an anti-amyloid monoclonal antibody approved by the FDA on January 6, 2023, also under an accelerated pathway based on removal of amyloid plaques. A prior version of this report included a review of donanemab. Due to the manufacturer receiving a Complete Response Letter from the FDA on January 19, 2023 for donanemab's accelerated approval biologics license application, we have removed donanemab from the report. Lecanemab was evaluated in a Phase Ill randomized clinical trial, CLARITY AD. The trial randomized 1,795 participants with early AD (i.e., mild cognitive impairment [MCI] or mild dementia due to AD) to a biweekly 10 mg/kg intravenous infusion of lecanemab or placebo. The primary clinical outcome was change in mean score on the Clinical Dementia Rating Scale - Sum of Boxes (CDR-SB). At 18 months, the lecanemab-treated group showed a statistically significant 27% slowing of cognitive decline compared with placebo, representing an average difference of about 0.5 points on the 18-point CDR-SB scale. Analyses of secondary endpoints, including other cognitive measures and patient and caregiver quality of life consistently favored the lecanemab-treated group. Among participants treated with lecanemab, 21.5% experienced amyloid related imaging abnormalities with edema/effusion (ARIA-E), ARIA-hemorrhage or superficial siderosis (ARIA-H), or both compared with 9.5% in the placebo group, and 3.5% of patients in the lecanemab group experienced symptomatic ARIA-E or -H compared with 0.2% in the placebo group. We remain uncertain that amyloid removal is an appropriate surrogate outcome for clinical benefit and instead look to the clinical outcomes found in randomized trials. However, there is ©Institute for Clinical and Economic Review, 2023 Page ES1 Alzheimer's Disease - Final Evidence Report Return to Table of Contents disagreement among experts about the clinical meaningfulness of the magnitude of change in CDR- SB in the lecanemab trial." We also remain concerned that real world ARIA occurrences and consequences may be more severe if, as expected, monitoring MRIs are not as frequent as in the clinical trial, the patient population treated differs from the trial population, and clinicians are less expert than those who participated in the randomized trial. In aggregate, the net health benefits of lecanemab in patients with early AD may be small or even substantial, but there remains a possibility of net harm from ARIA. We rate treatment with lecanemab in MCI due to AD or mild AD as "Promising but Inconclusive" (P/I). Table ES1. Evidence Ratings Treatment | Comparator | Evidence Rating Patients with Early Alzheimer's Disease Lecanemab | Supportive care only | P/l We estimated the lifetime cost effectiveness of lecanemab in addition to supportive care as compared to supportive care alone from a health care sector perspective (i.e., focusing on the direct medical care costs and health outcomes of the patient) and a modified societal perspective (i.e., including patient productivity impacts, caregiver time spent caregiving, caregiver quality of life, and caregiver direct medical costs). From both perspectives, lecanemab's annual price of $26,500 exceeded commonly used cost-effectiveness thresholds. ICER's Health Benefit Price Benchmark (HBPB) for lecanemab is $8,900 to $21,500, requiring a 66% to 19% discount from lecanemab's wholesale acquisition cost (WAC). Assuming lecanemab's current wholesale acquisition cost, approximately 5% of the roughly 1.4 million US patients eligible for AD treatment that targets beta-amyloid could be treated within five years without crossing the Institute for Clinical and Economic Review potential budget impact threshold of $777 million per year. Therefore, at current pricing and with projected uptake that is likely to exceed 5% of the eligible population, lecanemab's short-term potential budget impact exceeds our threshold. Additional efforts at achieving affordability and access must be considered, thus we are issuing an access and affordability alert. Appraisal committee votes on questions of comparative effectiveness and value, along with policy recommendations regarding pricing, access, and future research are included in the Report. Several key themes are highlighted below: o All stakeholders have a responsibility and an important role to play in ensuring that new treatment options for patients with AD are introduced in a way that addresses the impact on health inequities. This includes, for example, fair pricing according to value assessments from independent analysts to avoid financial toxicity, coverage of PET scans and ApoE €4 ©lnstitute for Clinical and Economic Review, 2023 Page ES2 Alzheimer's Disease - Final Evidence Report Return to Table of Contents genotype testing, and increasing capacity for screening, diagnosis, and treatment, particularly in rural or underserved areas. e Manufacturers of future treatments for Alzheimer's disease should follow the example set by the manufacturer of lecanemab by sharing a transparent, explicit justification for their pricing. However, pricing should be based on independent - not industry-funded - assessments of the drug. e Manufacturers should release all patient-level data in order to help patients, clinicians, researchers, and regulators to understand more about the link between amyloid reduction and cognitive outcomes. Failure to do so will impair scientific advances about the mechanisms underlying AD. ©lnstitute for Clinical and Economic Review, 2023 Page ES3 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 1. Background ICER reviewed aducanumab for early Alzheimer's disease in 2021 (Aducanumab for Alzheimer's Disease: Effectiveness and Value®). Much of the background information in this evidence report is updated from that report with additional contextual information about lecanemab. Alzheimer's disease (AD) is a fatal degenerative brain disease characterized by progressive loss of cognitive skills such as such as memory, language, navigation, and problem-solving, and physical function. It is the most common cause of dementia in the United States (US), accounting for up to 60-80% of all dementia diagnoses and is the sixth leading cause of death.® Direct costs of health care related to AD are estimated to be around $345 billion in 2023, and are projected to increase to just under $1 trillion in 2050.2 However, the economic burden of the disease may be underestimated, as many non-medical costs such as home safety modifications, adult day care services, and adverse effects on caregiver health and productivity may not be included in cost estimates.' AD affects an estimated 6.7 million Americans, or around 10% of the population aged 65 year or older, and around one-third of people aged 85 and older.? Early-onset AD, defined as the onset of AD prior to age 65, accounts for around 5% of cases and is associated with a larger genetic predisposition and differences in clinical and pathologic presentation.'® Two-thirds of those diagnosed with AD are women,? and there is evidence that symptoms of the disease may manifest differently in women and men, particularly with respect to neuropsychiatric symptoms.1? There are racial and ethnic differences in the incidence and prevalence of AD, with higher rates noted in the African American and Hispanic populations compared with non-Hispanic White and Asian populations.®13 The hallmark of AD is the progressive accumulation of beta-amyloid protein plaques and neurofibrillary tangles of hyperphosphorylated tau protein in the brain; these are hypothesized to lead to damage and eventual death of neurons over decades. Single-gene mutations that impact beta-amyloid formation (e.g., amyloid precursor protein and presenilin) are associated with early- onset AD. Genetic variants such as the apolipoprotein E (ApoE) £4 allele increase one's risk of developing late-onset AD; having one copy of the gene is associated with a two to threefold increase in developing AD, while two copies of the gene may increase risk of AD by as much as 15 times.!* Additionally, there are different forms of amyloid such as plaques, oligomers, and monomers, and the roles of these different forms and how specifically they are pathophysiologically associated with AD is not well understood; however, the neuronal damage from accumulation of amyloid is thought to trigger a cascade leading to impairment in cognitive domains such as memory, language, executive function, and visuospatial function, which results in the loss of ability to perform instrumental and basic activities of daily living (e.g., paying bills, bathing, dressing, etc.).' ©lnstitute for Clinical and Economic Review, 2023 Page 1 Alzheimer's Disease - Final Evidence Report Return to Table of Contents The course of AD can be described in three phases: preclinical disease, mild cognitive impairment (MCI) due to AD, and dementia due to AD. People living with AD begin to accumulate beta-amyloid in the brain in the preclinical phase up to 15 years prior to the onset of symptoms.2® These changes can be detected through the cerebrospinal fluid (CSF) (e.g., decreased beta-amyloid and increased CSF tau protein levels) and imaging (e.g., amyloid on positron emission tomography [PET] scans). CSF and imaging biomarkers can be used to identify AD pathology as well as to guide therapy and to monitor the impact of therapeutic interventions.17-1° Difficulty remembering recent conversations, names or events is often an early symptom, along with decreased awareness of financial scams?'; depression and apathy may also be seen at this stage. The diagnhosis of MCl is marked by a reduction in cognitive function; at this point, the patient can still live and function independently, although they typically show impairment in instrumental activities of daily living. Individuals are diagnosed with Alzheimer's dementia when there is impairment of two cognitive domains and these deficits significantly interfere with the ability of the patient to function independently at work or at home, although other forms of dementia may be present as well.2!. At this stage, a person typically needs a caregiver. As the disease progresses, people living with AD become less and less independent and the need for caregiving increases. Symptoms can include difficulty with communication, disorientation to time and place, confusion, poor judgment, and behavioral changes. At late stages, people living with AD may have difficulty speaking, swallowing, and walking. Eventually, many people living with AD require around-the-clock in-home or institutional care. More than 11 million family members and other caregivers and other persons provided an estimated 16 billion hours of unpaid care to people with AD or other dementias, putting these caregivers at risk for negative mental, physical, and emotional outcomes.? The average life expectancy for people with AD depends on multiple factors including age, functional status at diagnosis, and comorbidities. Estimates range from 4 to 8 years, but some people with AD live as many as 20 years after diagnosis.' Treatment of AD remains largely supportive, including creation and implementation of individualized care plans (e.g., treatment of dementia symptoms, medication and home safety assessments, advance care planning), caregiver education and support, care navigation, care coordination, and referral to community-based organizations for services and supports (e.g., adult day care, caregiver training, etc.).?2 Non-pharmacologic treatments include physical activity, which some studies have suggested may prevent or mitigate AD?>?* as well as behavioral and environmental strategies to ameliorate neuropsychiatric symptoms (e.g., agitation, delusions, disinhibition), and problem behaviors (e.g., resistance to care, hoarding, obsessive-compulsive behaviors).? Pharmacological therapy of AD has until recently focused mainly on symptom management. The most commonly prescribed drugs are the cholinesterase inhibitors (AChEI), including donepetzil, rivastigmine, and galantamine, and memantine, a drug that affects glutamine transmission. ©lnstitute for Clinical and Economic Review, 2023 Page 2 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Cholinesterase inhibitors are indicated in mild, moderate, and severe AD, while memantine is approved for moderate-to-severe AD. These drugs, either alone or in combination, are often used to treat the cognitive and functional symptoms of the disease, despite limited evidence of efficacy and significant side effects.?6%' Given the large and growing population of people with AD and the economic and human burden of AD, there is a tremendous need for disease-modifying drugs (i.e., drugs that slow or stop progression of AD). To date, more than 20 drugs targeting purported molecular pathways of AD (e.g., beta-amyloid or tau proteins} have either failed in clinical trials or are still in development. Aducanumab {(Aduhelm™, Biogen), a human monoclonal antibody, was the first putative disease- modifying drug to obtain accelerated approval from the FDA in June 2021, based on PET- documented removal of amyloid from the brain. Since then, two additional anti-amyloid monoclonal antibodies have submitted applications for accelerated approval for the treatment of AD: lecanemab (Legembi™, Eisai Co., Ltd) and donanemab (Eli Lilly & Co.). Lecanemab was approved on January 6, 2023 under accelerated approval and is the focus of this review. The manufacturer of donanemab received a Complete Response Letter for accelerated approval from the FDA on January 19, 2023. Therefore, although a prior version of the report contained a review of donanemab, we have removed it from this version. Table 1.1. Interventions of Interest Intervention Mechanism of Action D:'I:::y Prescribing Information Monoclonal antibody that binds to . «TM Leqembi™ (lecanemab) beta-amyloid protofibrils Intravenous 10 mg/kg biweekly mg: milligram, kg: kilogram ©lnstitute for Clinical and Economic Review, 2023 Page 3 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 2. Patient and Caregiver Perspectives ICER engaged with people living with Alzheimer's disease (AD) and caregivers, representatives from advocacy organizations, and clinical experts to understand the specific challenges associated with caring for persons with AD. We spoke with a total of 13 people with AD from across the US and across the disease spectrum (mild to severe), and in various living arrangements, and five caregivers from across the US with varied caregiving experiences (from early to late-stage AD). These were informal interviews; demographics of the sample are described further in the Supplement. We also drew on conversations from our 2021 Alzheimer's disease review.?® Individuals we spoke with emphasized the following issues, which are discussed in detail below: challenges with diagnosis, experience of coping with the diagnhosis and a new way of living, impact on caregiver quality of life, treatment concerns and goals, and financial impacts and disparities. Challenges with Diagnosis Initial symptoms of AD may start many years prior to receiving a diagnosis starting with gradual cognitive decline? including difficulties with memory and executive function (a term that describes abilities of planning, flexible thinking, and focusing attention).3%3! People with AD described forgetting meetings or names, getting lost, and feeling that something was "wrong"; sometimes they also received feedback from others who had noticed changes in behaviors. Caregivers also described that they began noticing subtle changes in their loved one's thinking and behavior, although sometimes identifying those changes was challenging if they did not live with the person with AD. Although an estimated 10% of people over the age of 65 are living with dementia due to AD, diagnosis Is often missed or delayed. People with AD described the diagnosis process as long and, for some, complicated by confusion over whether their symptoms were due to AD or another medical condition. When receiving the diagnosis, some people with AD reported having a negative experience with their physician, such as being told to "get their affairs in order" or that they would decline rapidly. On the other hand, one person with AD reported that their physician told them to focus on living life fully which allowed them to continue to focus on activities they enjoyed (e.g., dancing and exercise). After the diagnosis, lack of education about how to navigate the disease was described by both people with AD and caregivers. One person noted that "there needs to be more support when diagnosed. It is a complex process, especially when having other medical conditions to deal with." People living with AD reported wanting more information on such things as the disease and its course, clinical trials, and information on how the diagnosis can impact caregivers. Comprehensive care planning (e.g., functional assessment, review of current medications for high-risk medications, evaluation of home safety, caregiver needs, etc.), linkage to social services, management of comorbidities, and information on end-of-life care was also lacking in many cases. Finally, ©Ilnstitute for Clinical and Economic Review, 2023 Page 4 Alzheimer's Disease - Final Evidence Report Return to Table of Contents caregivers also reported on the lack of resources for them and having to educate themselves on AD to understand the best care for their loved one. Experience of Coping with the Diagnosis and a New Way of Living People living with AD described significant life changes after receiving the diagnosis. One person stated that their diagnosis "changed everything". Adjustments to daily life described to us included accepting more direction, using their phone to help with reminders, wearing headphones in noisy environments, moving to be closer to their family, and online grocery shopping. People with AD reported experiences of loss after the diagnosis, such as giving up certain activities, leaving their jobs, rehoming their animals, and confronting fears about loss of identity and the future loss of their independence. Depression, including suicidal thoughts, can also occur, and quality of life suffers even before severe memory deficit occurs. One person described that their "life has gone from a big platform to a small one." Coping mechanisms for the new changes in their lives included trying to maintain a positive attitude and staying busy, active, and social. AD also had an impact on relationships. People with AD described a loss of ability to be the type of relative they wished to be, such as wanting to look after their grandchildren, remembering special dates or birthdays, or developing new relationships. Having AD is financially challenging, both because people with AD may have to stop working and because of challenges with navigating the system when applying for Social Security Disability Insurance (SSDI) or moving onto Medicare. Caregivers described taking control of their loved one's finances early and the challenges of getting access to services. Some caregivers reported that they were unable to gain access to all services and had to use their own funds to provide support for later stages (e.g., 24-hour in-home support to allow caregivers to continue to work, medications, clothes, or hobbies), which could also cause changes their financial circumstances. Impact on Caregiver Quality of Life The impact of AD on caregivers is substantial. Nearly half of all caregivers who provide care to older adults do so for someone with dementia - often without training. Women are not only more likely to be caregivers but also to spend more time providing care than men. Surveys of caregivers show that they spend 40 to 60 hours per week directly caring for the patient; hours vary with severity of disease and care setting.* Caregivers described that their caregiving responsibilities evolved over time and with stage of disease, and ranged from paying the bills and driving, and as the disease progresses to moderate-to-severe dementia and the patient loses function, assisting with activities of daily living such as bathing and dressing. Some caregivers were able to obtain support from other family members and elder services programs to help keep their loved ones living at home; others transitioned their loved ones to assisted living or nursing homes. ©lnstitute for Clinical and Economic Review, 2023 Page 5 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Caregivers described the emotional toll of caregiving, reporting that it was difficult to witness changes in how their loved one interacts with and experiences the world. One caregiver reported that it "took a while to realize that although part of her is still "mom", not everything that comes out of her mouth is really her... Those who aren't in the caregiving situation find it hard to understand where their heads are. [We may] come across as too pragmatic that we seem cold, but Surveys have quantified the toll on caregivers - as individuals moved from mild to severe AD, the financial, physical, psychosocial, social, and personal strain as measured by the Modified Caregiver Strain Index increased from an average score of 9.0 to 17.5 {(out of a maximum of 26), indicating a substantial increase in caregiver impact.* As a result, caregivers often suffer physical and mental health consequences including increased chronic health conditions, depression and isolation, and increased use of the health care system. Several caregivers reported that there is not enough support available for their emotional I" we all need to function, and this takes an incredible tol and mental health needs and that there is variation in the degree of support available across the US for both patients and their caregivers. When asked about their broader experience of caregiving, some caregivers describe the impact on their own life plans, including the ability to work and personal relationships. Caregivers described that because of their caregiving responsibilities, they have had less time for their own family, such as fewer vacations or leisure activities. Some caregivers reported resigning from their jobs to be a caregiver, as they were unable to manage both caregiving and job responsibilities. Another caregiver reported that they had to spend time away from their family to care for their loved one with AD which caused strain on their relationship at home. One caregiver described that "I feel like | haven't had the chance to live my life" and another caregiver told us that "[Person with AD] has often asked me if | am going to have kids... that is the only thing that may have stopped." Treatment Concerns and Hopes People with AD and caregivers would like more options for treatments. The main goal of treatment is not to cure, but to slow or halt the progression of the disease and maintain their current level of functioning (e.g., independence, hobbies, personal care) as long as possible. One patient described that "more time doesn't do it for me. | would say more quality time". In terms of anti-amyloid therapies, both people living with AD and caregivers were interested in any treatment that would help slow disease progression. However, both groups described concerns about the side effect of brain swelling and questioned whether the gains in quality of life that might be seen with treatment outweighed that risk. There were also logistical concerns about a treatment with regular infusions, particularly for those persons who do not drive, and concerns about an increased burden on caregivers. Caregivers also reported treatment accessibility issues due to the location of medical centers, the limited number of physicians and health systems offering treatment and concerns about insurance coverage [in the context of aducanumab], and concern over their loved one experiencing pain or anxiety from the infusion. ©lnstitute for Clinical and Economic Review, 2023 Page 6 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Disparities in Clinical Care and Research Both people with AD and caregivers raised concerns surrounding disparities in clinical care and research. For example, although African Americans make up only around 9% of the US population, they represent 13.8% of persons with dementia, with Black women having the highest prevalence. Black persons are more likely to experience neuropsychiatric symptoms, behavioral changes such as agitation and aggression, abnormal sleep, and motor disturbances.3? Furthermore, Black persons are more likely to be diagnosed with dementia at a later stage.3? There are also inequities in caregiving, with Black and Hispanic caregivers more likely to provide a high intensity level of care, and Black caregivers more likely to be the sole unpaid caregiver for their care recipient and less likely to receive respite services and information from medical care providers.333* Both people with AD and caregivers raised concerns about the limited or culturally misaligned information provided to communities of color about AD or how to provide care for loved ones with AD, and believed that more resource sharing and more support group leaders who are people of color could raise awareness in the community. In particular, increasing awareness about the Improving HOPE for Alzheimer's Act of 2020,% which aims to increase use of comprehensive care planning through education of clinicians, patients, and caregivers, could improve access to care. Broader access to effective therapies and increasing diversity among providers for dementia care could also help address disparities in treatment. People living with AD noted that "everyone should have the right to participate in clinical trials" and those who participated felt fortunate to live close to the large academic centers where trials are being conducted. However, people of color are less likely to be able to participate due to having other medical comorbidities, location of the clinical trials, or not having a care partner. Patient groups pointed to a need to engage, recruit, and retain diverse populations in Alzheimer's research and clinical trials to help decrease health inequities. ©lnstitute for Clinical and Economic Review, 2023 Page 7 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 3. Comparative Clinical Effectiveness 3.1. Methods Overview Scope of Review We reviewed the clinical effectiveness of lecanemab, in addition to supportive care, versus supportive care alone for the treatment of individuals with early Alzheimer's Disease (AD) (i.e., MCI due to AD and mild dementia due to AD) with evidence of AD pathology. We sought and reviewed evidence, where available, on patient-important outcomes, including change in the ability to maintain independence and cognitive function and perform activities of daily living, as well as delaying entry into institutional care, quality of life, caregiver impact, and behavioral change. We also reviewed evidence on changes in biomarkers (e.g., level of beta-amyloid, level of tau and phosphorylated tau [p-tau], and brain volume). We also aimed to evaluate updated evidence for aducanumab that was not covered in the previous August 2021 review.? We found no new clinical evidence for aducanumab that would have affected our assessment and thus it is not reviewed in this report. The full scope and procedures for the systematic literature review are detailed in the supplement. Evidence Base Lecanemab Evidence informing our review of lecanemab was derived from one Phase Il randomized controlled trial (RCT): CLARITY AD. We also reviewed the lecanemab Phase Il trial (G000-201), but because there were differences in the trial objectives, dosing, and design, it was not a primary focus of our review. G000-201 is described in Section D of the Supplement. CLARITY AD was a Phase Il placebo-controlled RCT that evaluated the efficacy of 10 mg/kg of intravenous lecanemab administered every two weeks versus placebo for 18 months, with an open- label extension phase with all participants receiving lecanemab for an additional 27 months.3® The primary outcome was change from baseline in CDR-SB at 18 months. The open-label extension phase examined treatment-emergent adverse events and change in CDR-SB. Participants were included if they were aged between 50 and 90 years and had a mild cognitive impairment due to Alzheimer's disease or mild Alzheimer's disease dementia, amyloid positivity as determined by PET or CSF (AP1-42), objective impairment in episodic memory, had a Mini-Mental State Exam (MMSE) score of 22-30, BMI 17-35, and, if they were receiving any treatment for AD symptoms, were on a stable dose for 12 weeks prior to baseline. Full inclusion and exclusion criteria are described in Table D3.1. in the Supplement. Baseline characteristics and cognitive scores for CLARITY AD are outlined in Table 3.1. ©Institute for Clinical and Economic Review, 2023 Page 8 Alzheimer's Disease - Final Evidence Report Return to Table of Contents ARIA was assessed by MRI at baseline and five additional time points across the study and presence of ARIA could have led to the study drug being held or discontinued. Table. 3.1. Baseline Characteristics and Cognitive Measures for CLARITY AD?6 . e - N=1795 Baseline Characteristics and Cognitive Scores (lecanemab, N=859, placebo, N=875*) Age in Years, Mean (SD) 71.2(7.9) Sex, Female % 52.3% Race/Ethnicity 76.9% White, 2.5% Black, 17% Asian, 3.7% Other, 12.4% Hispanic Concomitant AD medication 52.8% AChEI or memantine ApoE &4 Carrier 68.8% ApoE €4 Heterozygote 53.3% ApoE €4 Homozygote 15.5% Stage of Disease 61.9% MCI due to AD and 38.1% Mild AD MMSE, Mean (SD) (Range) 25.6 (2.2) (22-30) CDR-SB, Mean (SD) {Range) 3.2 (1.34) (0.5-8.5) ADAS-Cogl4, Mean (SD) (Range) 24.4 (7.32) (4.7-60.7) ADCS-MCI-ADL, Mean (SD) (Range) 41.1(6.75) (12-53) ADCOMS, Mean {SD) (Range) 0.399 (0.147) (0.07-0.94) Baseline Beta-amyloid, Mean Centiloids (SD) 76.48 (43.3)1 AChEI: acetylcholinesterase inhibitor, AD: Alzheimer's disease, ADAS-Cog 14: Alzheimer's Disease Assessment Scale-Cognitive Subscale 14-item, ADCOMS: AD composite score, ADCS MCI-ADL: Alzheimer's Disease Cooperative Study - Activities of Daily Living Inventory {(MCl version), ApoE: apolipoprotein E, CDR-SB: Clinical Dementia Rating scale Sum of Boxes, MMSE: Mini-Mental State Examination, SD: standard deviation *mITT population tIncluded only participants enrolled in the PET substudy {(N=354 lecanemab and N=344 placebo) OInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page 9 Return to Table of Contents 3.2. Results Clinical Benefits In this report, we describe the change in cognitive function, health-related quality of life, beta- amyloid, and other biomarkers (tau, p-tau, and brain volume) from baseline to 18 months after treatment initiation for lecanemab. A variety of cognitive measures were used across the trials. In Section Al of the Supplement, we provide definitions of each of the cognitive outcomes. Table 3.2. provides minimal clinically importance differences (MCID) in the published literature associated with cognitive outcomes used across the trials, although there is not universal agreement on what constitutes a clinically relevant change for persons with AD. While MCID applies to changes for a specific patient, when averaged in aggregate it can provide context for the magnitude of changes overall. Itis important to remember, though, that if patient responses are normally distributed, half of patients will have outcomes better than the average and half worse than the average. We discuss results for lecanemab in the context of MCIDs to help contextualize whether statistically significant results may also be clinically relevant, as statistical significance can be a reflection of sample size, a clinically relevant change, or both. Table 3.2. Minimal Clinically Importance Differences for Cognitive Outcomes Cognitive Outcome Minimal Clinically Important Difference (MCID) Clinical Dementia Rating - Sum of Boxes (CDR-SB) Change of 0.98-1.63 df:rrn':enni:ade to AD and mild AD Alzheimer's Disease Assessment Scale - Cognitive Subscale (ADAS-Cog), including ADAS-Cog-13 and ADAS-Cogl4. Change of 2 points for MCI due to AD* and 23 points for mild AD.3940 Alzheimer's Disease Cooperative Study - Activities of Daily Living Inventory (MCI version) (ADCS-MCI-ADL) There are no data on MCID. AD Composite Score (ADCOMS) There are no data on MCID. AD: Alzheimer's disease, MCl: Mild cognitive impairment Lecanemab Cognitive outcomes The primary outcome in CLARITY AD trial was the change from baseline in CDR-SB score at 18 months. Participants treated with lecanemab had 27% less decline in CDR-SB compared to placebo at 18 months, with a mean difference of -0.45; 95% Cl: -0.67 to -0.23; p<0.0013¢ (Table 3.3.). This mean difference in CDR-SB was less than what some consider as MCID for the scale.3"*! Changes in CDR-SB remained fairly consistent across the sensitivity analyses that imputed data, included all randomized participants, or censored assessments after ARIA-E had occurred (Supplement Table D3.22).% ©lnstitute for Clinical and Economic Review, 2023 Page 10 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Other measures of cognition assessed in the trial also showed statistically significant differences favoring the lecanemab treated group (Table 3.3.). Participants treated with lecanemab had less decline in ADAS-Cog14 compared to placebo at 18 months, with a mean difference of -1.44; 95% Cl: -2.27 to -0.61; p<0.0013, less than the 2-3 point change considered as the MCID for this scale for people with MCl and mild AD.3%4° Those treated with lecanemab also had less decline in ADCS-MCI- ADL compared to placebo, with a mean difference of 2.0; 95% Cl: 1.2 to 2.8; p<0.001.36 and less decline in ADCOMS compared to placebo, with a mean difference of -0.05; 95% Cl: -0.074 to -0.027; p<0.001.3% This mean difference for ADCOMS was the same as that reported in the Phase Il GOOO- 201 trial. Health-Related Quality of Life Definitions of health-related quality of life scales are provided in Section A of the supplement. Changes in health-related quality of life measures were available for around 85% of participants at 18 months. For participant-reported outcomes, participants in the lecanemab group had 49% less decline on European Quality of Life - 5 dimensions (5 level version) (EQ-5D-5L) at 18 months compared to placebo (p<0.01). There was also 56% less decline on the Quality of Life in Alzheimer's Disease (QOL-AD) in the lecanemab group compared to the placebo group (p<0.01).*? For caregiver-reported outcomes, caregivers of participants in the lecanemab group reported 38% less decline on Zarit Burden Interview at 18 months compared to placebo {p<0.001) and 23% less decline on the QOL-AD (subject by proxy) compared to placebo (p<0.05).4? See Supplement Table D3.4. Beta-Amyloid Levels Changes in beta-amyloid, as measured in centiloids using florbetaben, florbetapir, or flutemetamol PET tracers, were assessed in a substudy including 698 participants (n=354 in lecanemab and n=344 in placebo). At 18 months, participants in the lecanemab group had a larger amount of beta- amyloid removed compared with the placebo group (mean difference -59.12; 95% Cl: -62.64 to - 55.60; p<0.001),%¢ corresponding to a calculated mean percentage difference versus placebo of - 76.0%. Additionally, at 18 months, 32.4% of those in the lecanemab group were amyloid negative (defined as having <30 centiloids on PET scan) compared to 7.8% in the placebo group.** CSF and plasma markers of AB42/40 were also consistent with amyloid removal in the lecanemab group. Descriptions of assessment and findings for Plasma and CSF measures of amyloid are in provided in Section D of the Supplement. Other Biomarker Outcomes There were statistically significant reductions in several tau biomarkers in the lecanemab treated group (CSF t-tau and CSF and plasma p-tau) in the lecanemab group, compared with the placebo group at 18 months3® (Supplement Table D3.4). In terms of brain volume, there was greater decrease in whole brain volume and a greater increase in ventricular volume in the lecanemab ©lnstitute for Clinical and Economic Review, 2023 Page 11 Alzheimer's Disease - Final Evidence Report Return to Table of Contents group compared to the placebo group at month 18. There was less atrophy in hippocampal volume in lecanemab as compared to placebo.®® At the time of this draft report, the investigators note that tau PET and volumetric MRI results have not been fully analyzed®® and thus interpretation of these outcomes are limited. Detailed descriptions of these assessments and results of other biomarkers are in provided in Section D of the Supplement. Table 3.3. Key Trial Results Measure CLARITY AD* Lecanemab Placebo Baseline (SD) (Range) 3.17 (1.34) (0.5 to 8.0) 3.22(1.34) (0.5 t08.5) Timepoint 18 months CDR-SB N 714 757 Mean Change 1.21 1.66 Difference in Mean Change {95% Cl) | -0.45 (-0.67 to -0.23) REF Baseline (SD) (Range) 24.45 (7.08) (4.7 to 47.7) 24.37 (7.56) (5.0 to 60.7) Timepoint 18 months ADAS-Cogls N 703 738 Mean Change 4.14 5.58 Difference in Mean Change (95% Cl) | -1.44 (-2.27 to -0.61) REF Baseline (SD) (Range) 0.398 (0.147) (0.08 t0 0.94) | 0.400 (0.147) (0.07 to 0.91) Timepoint 18 months ADCOMS N 708 749 Mean Change 0.164 0.214 Difference in Mean Change {95% Cl) | -0.05 (-0.074 to -0.027) REF Baseline (SD) (Range) 41.2 (6.6) (13 to 53) 40.9 (6.9) (12 to 53 Timepoint 18 months :gfs'Mc" N 676 707 Mean Change -3.5 -5.5 Difference in Mean Change {95% Cl) | 2.0(1.2 to 2.8) REF ADAS-Cog 14: Alzheimer's Disease Assessment Scale-Cognitive Subscale 14-item, ADCOMS: AD composite score, ADCS MCI-ADL: Alzheimer's Disease Cooperative Study - Activities of Daily Living Inventory (MCI version), CDR-SB: Clinical Dementia Rating scale Sum of Boxes, SD: standard deviation Harms Lecanemab Total adverse events and serious adverse events were slightly higher in the lecanemab versus placebo group (Table 3.4.) in the CLARITY AD trial. There were 6 deaths in lecanemab treated group and 7 deaths in placebo group, with myocardial infarction, cerebral macrohemorrhage, cerebrovascular accident, respiratory failure, and COVID-19 listed among the causes of death in both groups.3® There were no deaths directly attributed to ARIA in either group. Discontinuation in CLARITY AD was lower than the Phase Il (G0O00-201) trial, and fairly similar across groups (lecanemab: 18.8% versus placebo: 15.6%). The most common reasons for discontinuation were ®Institute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page 12 Return to Table of Contents adverse events, withdrawing consent, and choosing to discontinue from treatment regimen. Discontinuation of treatment due to adverse events were higher in lecanemab (6.9% vs. 2.9%) as were discontinuation of trial due to adverse events (5.7% vs. 3.1%) (Table 3.4). Participants in the lecanemab group also reported more infusion-related reactions compared to the placebo group (26.4% vs 7.4%). 36 ARIA due to edema or effusion (ARIA-E) or brain microhemorrhage or localized superficial siderosis reflecting prior hemorrhage (ARIA-H) were of interest to the review. A total of 21.5% participants in the lecanemab group experienced either ARIA-E or ARIA-H, compared to 9.5% in the placebo group (Table 3.4., Supplement Table D3.11). As expected, ARIA-E occurred more often in the lecanemab group compared with placebo (12.6% vs 1.7%), and 2.8% of participants in the lecanemab group experienced symptomatic ARIA-E (e.g., headache, visual disturbance, or confusion or severe radiographic changes on MRI). ARIA-E occurred more frequently in ApoE €4 carriers compared to non-carriers in the lecanemab treated group (3.4% vs 1.4%), and also occurred more in ApoE €4 homozygotes compared to heterozygotes (9.2% vs 1.7%).3% ARIA-E generally occurred in the first three months of treatment (71%), was mostly mild-moderate in severity, and the investigators reported that 81% of ARIA-E resolved within four months.** Seven participants (0.8%) in the lecanemab group had a significant adverse event (SAE) associated with ARIA-E. Recurrent ARIA-E occurred more often in the lecanemab group compared with the placebo group (3.1% vs 0.1%, Supplement Table D3.11). ARIA-H, which was mainly due to cerebral microhemorrhage, occurred more frequently in the lecanemab group compared with placebo (17.3% versus 9.0%); very few cases of ARIA-H were symptomatic (0.7% in lecanemab and 0.2% in placebo). ARIA-H was more likely to co-occur with ARIA-E in the lecanemab group (8.2%) compared to placebo (1.0%).3¢ Like ARIA-E, ARIA-H occurred more frequently in ApoE &4 carriers compared with non-carriers (19.7% versus 11.9% in the lecanemab treated group) and was more frequent in ApoE €4 homozygotes compared to heterozygotes (39% versus 14%).3% Unlike ARIA-E, which occurred early in the study period, ARIA-H occurred randomly across the study period.** There was one reported case of cerebral macrohemorrhage that occurred 30 days after stopping the study drug and thus its relationship to lecanemab treatment is uncertain. Full data are not available from the open-label extension (OLE) phase of CLARITY AD. During this phase, safety assessments were conducted every 6 months. The safety MRI schedule initially followed the same schedule as the core phase for 6 months (weeks 9, 13, and 6 months)} and reduced to every 6 months thereafter. At the time of posting this report, three deaths have been reported in the lecanemab group due to cerebral macrohemorrhage (0.1%). One participant was reportedly on the anticoagulant apixaban and one participant was given tissue plasminogen activator (tPA) while being treated for an acute stroke.**® The third participant received lecanemab during the OLE and had what appeared to be severe symptomatic ARIA-E and ARIA-H ©lnstitute for Clinical and Economic Review, 2023 Page 13 Alzheimer's Disease - Final Evidence Report Return to Table of Contents with cognitive dysfunction and seizures, and died days later.*' The trial sponsor reports that they are still investigating whether this death is related to lecanemab treatment.' Table 3.4. Adverse Events Harms CLARITY AD3 Lecanemab Placebo Any Adverse Event, n/N (%) 798/898 (88.9%) 735/897 (81.9%) Serious Adverse Event, n/N (%) 126/898 (14.0%) 101/897 (11.3%) Discontinuation Due to Adverse Event, n/N (%) | 62/898 (6.9%)* 26/897 (2.9%)* Any ARIA-H, n/N (%) 155/898 (17.3%) 81/897 (9.0%) Any ARIA-E, n/N (%) 113/898 (12.6%) 15/897 (1.7%) Symptomatic ARIA-E, n/N (%) 25/898 (2.8%) 0/897 (0%) ARIA-E: amyloid-related imaging abnormalities due to edema/effusion, ARIA-H: amyloid-related imaging abnormalities due to hemorrhage or superficial siderosis, N: total participants, n: number of participants *Per the protocol, primary reasons for any discontinuation from the study are: AE, lost to follow-up, subject choice, withdrawal of consent, pregnancy, study terminated by sponsor, and other. Subgroup Analyses and Heterogeneity CLARITY AD trial included a more racially and ethnically diverse sample than previously published Alzheimer's disease trials, see below for full evaluation of clinical trial diversity. In the CLARITY AD trial, over half the sample had two or more comorbidities (e.g., diabetes, hypertension, etc.) and 4.5% were on anticoagulants, more closely representing participants who may seek this therapy in real-world clinical settings. Prespecified subgroup analyses were conducted on the cognitive outcomes across various subgroups of interest. There were four subgroup analyses that appeared to have consistent results across the cognitive measures. The effect of lecanemab on cognitive outcomes appeared to vary across ApoE €4 carrier status, with lecanemab showing a larger effect in ApoE &4 noncarriers compared to carriers and having the least effect in ApoE ¢4 homozygote participants, particularly on the CDR-SB and ADCOMS measures. When disease state was examined, there was similar relative change in CDR-SB in MCI compared with mild AD (28% versus 27% change, respectively). There was a trend towards greater slowing of decline in older participants (275 years) compared to younger participants (<65 years) in 3 of the 4 cognitive outcomes. Finally, male participants showed a greater slowing of decline compared to female participants.*® See Figure 3.1. For CDR-SB specifically, the treatment effect remained consistent across participants with comorbidities, such as hypertension, diabetes, heart disease (Supplement Table D3.17).%? Detailed results on subgroups are presented in Supplement Tables D3.18-23 and additional subgroups are described in Supplement D2. Caution must be taken in the interpretation of these subgroup effects as they were likely underpowered and did not correct for multiplicity. OInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page 14 Return to Table of Contents Figure 3.1. Forest Plot of Subgroup Analyses for CDR-SB at 18 Months {Age, ApoE €4 Genotype, and Sex) Yo Adjusted mean Subgroup Slowing difference in CDR-SB Ama mn{amer ............. 4.1 ...................... 0 T.S ............................................................... ApoE4 heterozygote 30 0.5 I-.-| ApoE4 homozygote 22 0.28 I i | Female 12 02 -- Male 43 -0.73 -E- <65 Years 6 -0.08 -a- 65-74 Years 23 -0.37 -a- ==T5 Years 40 072 -B- ©lnstitute for Clinical and Economic Review, 2023 Page 15 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Evaluation of Clinical Trial Diversity Table 3.5. Sample Diversity Ratings on Race and Ethnicity, Sex, and Age (Older Adults) Race and - Trials Ethnicity (All Race and Ethnicity Sex Age Participants) (US subgroup) (Older adults) G000-201 (Phase Il) Poor NR Good NE CLARITY AD (Phase Ill) Fair Fair Good NE NE: Not Estimated, NR: Not Reported. We evaluated the demographic diversity of the clinical trials using the ICER-developed diversity rating tool. Table 3.5. presents sample diversity ratings on race and ethnicity, sex, and age (older adults) on the two key trials in our review. Details on each of the demographic categories are provided below. Additional details on the clinical trial diversity rating tool and detailed information on the scores and rating of each trial are provided in Supplement D1. Race and Ethnicity: The Phase Il trial did not report any data on race and ethnicity and thus was rated as "poor." The Phase lll CLARITY AD trial, which we rated as "fair" on racial and ethnic diversity, had an adequate representation of White and Asian individuals compared to the disease prevalence; however, Black or African American and Hispanic or Latino individuals were underrepresented. Because the CLARITY AD trial was a multinational study, we also separately evaluated the subpopulation of patients enrolled in the US. Compared to the disease prevalence, White and Hispanic individuals were adequately represented in the US subpopulation of the CLARITY AD trial, while Black or African American and Asian individuals were underrepresented. Thus, similar to the overall population, we rated the US subpopulation in the CLARITY AD trial as "fair" on racial and ethnic diversity. See Supplement Tables D1.9.-D1.11. Sex: Both trials adequately represented males and females; therefore, both trials were rated as llgood." Age: As Alzheimer's disease typically develops in older adults, we did not attempt to assess diversity in terms of age. Uncertainty and Controversies If the amyloid hypothesis is correct, it would be expected that removal of amyloid should be associated with treatment effect. For lecanemab we do not yet have adequate data showing such correlations (see Supplement Table D3.24.) or data showing differences in outcomes by achieving or not achieving amyloid negativity. Additionally, published systematic literature reviews that have incorporated both published and unpublished trials on anti-amyloid medications report that, while these medications were effective at reducing beta amyloid in the brain, the effect on slowing cognitive decline was inconsistent.**>! However, limitations of those reviews such as lack of full ©lnstitute for Clinical and Economic Review, 2023 Page 16 Alzheimer's Disease - Final Evidence Report Return to Table of Contents data availability and presentation of only aggregated data, prevents assessments of heterogeneity in response to amyloid reduction and slowing of cognitive decline.>> See Supplement D5 for summaries of these reviews. MCID refers to changes within a specific patient. Aggregate measures, such as mean change in an outcome, will potentially obscure changes in individual patients that are above or below the MCID. It would be helpful in assessing magnitude of benefit relative to clinical benefit if manufacturers provide analyses of the percentage of patients who experience decreases in an outcome measure beyond a prespecified endpoint (percentage clinically important decline, for instance) in the placebo and treatment groups. However, with or without such an analysis, unless changes in the assessed outcome measure are very non-normally distributed, comparison to MCID will be helpful to those assessing the data in understanding the magnitude of benefits for patients. There have been concerns among experts that improvements seen in trials of anti-amyloid therapies may reflect "functional unblinding". This could occur on subjective scales when patients, caregivers, and providers recognize that a patient who develops ARIA is very likely receiving an active treatment rather than placebo. As noted in our prior report, the best approach to this would be to ensure that there are equal numbers of patients managed as if they had developed ARIA in the placebo arm so as to maintain blinding. As this was not done, ICER reviewed data across trials (data not shown) to see if frequency of ARIA correlated with benefit. We did not find such a correlation and so continue to consider it relatively unlikely that the results seen in trials of anti- amyloid therapies are due to functional unblinding. The FDA accepted a statistically significant change in CDR-SB in the aducanumab trials as sufficient evidence of benefit to proceed with accelerated approval; however, experts disagree on this point.">® Some experts have suggested that the minimal clinically important difference (MCID) for CDR-SB is on the order of 1 or 2 points.3" Others suggest that MCID only reflects mean group differences and does not capture patient-level change and thus may not accurately reflect benefit to individual patients.>* Thus, the absolute difference in CDR-SB of 0.45 points between groups, while statistically significant, may or may not result in a change in status that is meaningful to individual patients and caregivers. A recent web-based survey to individuals at risk for or with AD reported that concepts that were meaningful to them went beyond cognition and functioning alone, and expanded to desire to maintain independence, emotional wellbeing, safety, and physical and mental health, which are currently not well captured in measures used in clinical trials of AD patients such as CDR-SB.>> In CLARITY-AD, assessments of patient and caregiver quality of life in CLARITY-AD do suggest measurable differences due to treatment with lecanemab; further details and longer-term follow-up are needed to fully understand the potential benefits of lecanemab treatment on this domain. Even in people with cognitive abnormalities and imaging showing amyloid, some individuals likely have other causes of dementia.>® A treatment that removes amyloid would have been anticipated ©lnstitute for Clinical and Economic Review, 2023 Page 17 Alzheimer's Disease - Final Evidence Report Return to Table of Contents to be most effective in patients least likely to have other forms of dementia. Patients who are older and male are at higher risk for vascular dementia, and APOE &4 is a known risk factor for AD. As such, older male participants without APOE €4 would be more likely to form a group in CLARITY-AD that has more vascular dementia (and possibly other non-Alzheimer's-type dementia). However, the actual subgroup analyses in CLARITY-AD show just the opposite (see Figure 3.1.). This s in contrast to the EMERGE trial for aducanumab, where similar changes in CDR-SB were seen overall, but where APOE &4 carriers had greater improvement than non-carriers. While a substantial percentage of participants treated with lecanemab reached amyloid negativity by PET scan, it should also be noted that 7.8% of participants in the placebo group were also amyloid negative at the end of the trial. This points to the complexity of the pathophysiology of AD and underscores that we do not fully understand the role of amyloid in AD and what factors (e.g., degree or rate of amyloid removal, a threshold of amyloid removal, or removal of certain amyloid subspecies) may impact clinical outcomes. The effectiveness of lecanemab in people with moderate or severe AD is unknown, as the drug was not studied in these groups. It is uncertain whether treatment should be continued indefinitely or discontinued either when a certain degree of amyloid clearance is attained or at a specific point in the course of the disease. ARIA is the major adverse effect seen with anti-amyloid therapies. Risks of ARIA in real world use may be greater than was seen in clinical trials given issues including that participants in clinical trials typically do better than randomly selected patients, clinical expertise is likely to be less in real world use, and labeling is unlikely to require the intensity of MRI monitoring that was done in CLARITY-AD. Furthermore, concerns about an increased risk of cerebral hemorrhage with concomitant use of anticoagulants have heightened with the report of three deaths in the lecanemab open label extension study in participants receiving anticoagulation. The average age of participants in CLARITY-AD was 71.2 years and included participants with some comorbidities. However, since two-thirds of individuals in the US with AD are 75 years old or older, who are also more likely to have significant comorbidities, the trial results may not be generalizable to the broader Alzheimer's population. ©lnstitute for Clinical and Economic Review, 2023 Page 18 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 3.3. Summary and Comment An explanation of the ICER Evidence Rating Matrix (Figure 3.2) is provided in the Supplement. Figure 3.2. ICER Evidence Rating Matrix Comparative Clinical Effectiveness High o Certainty Q S 3 > w Q = £ > Moderate -~ . £ Certainty o o Q T © > Q =~ Low " , Certainty | * " | > Negative Comparable Small Substantial Net Benefit Net Benefit Net Benefit Net Benefit Comparative Net Health Benefit A = "Superior" - High certainty of a substantial (moderate-large) net health benefit B = "Incremental" - High certainty of a small net health benefit C = "Comparable"- High certainty of a comparable net health benefit D= "Negative"- High certainty of an inferior net health benefit B+= "Incremental or Better" - Moderate certainty of a small or substantial net health benefit, with high certainty of at least a small net health benefit C+ = "Comparable or Incremental" - Moderate certainty of a comparable or small net health benefit, with high certainty of at least a comparable net health benefit C- = "Comparable or Inferior" - Moderate certainty that the net health benefit is either comparable or inferior with high certainty of at best a comparable net health benefit C++ = "Comparable or Better" - Moderate certainty of a comparable, small, or substantial net health benefit, with high certainty of at least a comparable net health benefit P/I = "Promising but Inconclusive" - Moderate certainty of a small or substantial net health benefit, small (but nonzero) likelihood of a negative net health benefit 1 = "Insufficient" - Any situation in which the level of certainty in the evidence is low The FDA approval of aducanumab, the first disease-modifying therapy for AD in June of 2021, based on biomarker evidence of beta-amyloid removal but with inconsistent clinical benefit from identically designed trials, brought forth several controversies, including how extensively amyloid is involved in the pathogenesis of AD and whether removal could result in meaningful clinical benefit, how to interpret outcomes from trials with protocol changes and that were stopped early for futility, and how great the risk of ARIA and its potentially severe consequences may be with use in the real world with less monitoring than clinical trials. Many prior anti-amyloid drugs have not ©lnstitute for Clinical and Economic Review, 2023 Page 19 Alzheimer's Disease - Final Evidence Report Return to Table of Contents demonstrated clinical benefits, and results reported subsequent to the approval of aducanumab were also negative for crenezumab and gantenerumab. Although it is possible to interpret these results as showing that inadequate amyloid was removed by those two therapies to achieve benefit, it is also possible that there is something specific to lecanemab's mechanism of action that resulted in cognitive benefits, alone or in combination with amyloid removal. We do not feel that evidence is adequate at this point to assume that clearance of amyloid will necessarily improve cognitive outcomes. Patient-level evidence across trials and agents showing that degree of amyloid removal correlates with clinical benefit would be an important step in establishing amyloid removal as an important surrogate outcome, although even then, surrogate endpoints may fail to predict the effect of treatment on a clinical outcome, either due to the complexity of disease pathophysiology or unintended effects on other outcomes."" Since removal of amyloid remains unproven as a surrogate outcome, it is important to look to clinical outcomes directly. Any cognitive benefits seen with an anti-amyloid therapy must be weighed against harms, and particularly the risk of ARIA. CLARITY AD, a Phase lll trial testing the efficacy of lecanemab, is the first clinical trial of an anti- amyloid antibody to clearly demonstrate that amyloid clearance is associated with a slowing of cognitive decline in participants with early AD. The 27% relative change in CDR-SB is similar to the change seen in EMERGE, the positive aducanumab trial. The CLARITY AD trial also showed positive results on other cognitive, MRI, and biomarker outcomes and on patient and caregiver quality of life, consistent with the primary outcome. However, whether the absolute change in CDR-SB, the primary outcome in the trial, is clinically relevant remains less certain. As a comparison, one RCT of donepezil 10 mg showed a statistically significant change in CDR-SB compared with placebo over 24 weeks of treatment (-0.60, p=0.0007).>® Additionally, as discussed above (see Figure 3.1 showing subgroup forest plots), we have some concerns that cognitive benefits were greatest in the subgroup of participants most likely to have included participants with other etiologies of dementia. As such, we have some concerns that lecanemab's mechanism of action may not be fully understood. Given all this, and the many prior negative trials of anti-amyloid therapies, we have some remaining uncertainties even about clinical benefits when looking at a single randomized trial. Rates of ARIA were less than seen in the aducanumab trials. Although there was no significant difference in the rate of death during the core clinical trial, there were three recently reported deaths in the open label extension trial, and thus we remain concerned that real world ARIA occurrences and consequences may be more severe if, as expected, monitoring MRIs are not as frequent as in the clinical trial, the patient population treated differs from the trial population, and clinicians are less expert than those who participated in the randomized trial. In aggregate, the net health benefits of lecanemab in participants with early AD may be small or even substantial, but there remains a possibility of net harm from ARIA. We rate treatment with lecanemab in MCI due to AD or mild AD as "Promising but Inconclusive" (P/1). ©lnstitute for Clinical and Economic Review, 2023 Page 20 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Table 3.6. Evidence Ratings Treatment | Comparator | Evidence Rating Patients with Early Alzheimer's Disease Lecanemab | Supportive care | P/l CTAF Votes Table 3.7. CTAF Votes on Comparative Clinical Effectiveness Questions Question Yes No Is the evidence adequate to demonstrate that the net health benefit of lecanemab 3 12 added to supportive care is superior to that provided by supportive care alone? A majority of the panel voted that the evidence is not adequate to demonstrate that the net health benefit of lecanemab added to supportive care is superior to supportive care alone. The panel expressed concerns around uncertainty around the amyloid hypothesis and the clinical meaningfulness of the results, as well as possible functional unblinding in the lecanemab trial, and potential serious adverse events. Voting panel members who voted that the evidence was adequate to demonstrate the net health benefit of lecanemab added to supportive care is superior to supportive care alone cited considering how devastating of a disease Alzheimer's is, along with the importance of letting patients decide if results were clinically meaningful in light of any risks. ©Institute for Clinical and Economic Review, 2023 Page 21 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 4. Long-Term Cost Effectiveness 4.1. Methods Overview The primary aim of this analysis was to estimate the lifetime cost effectiveness of lecanemab in addition to supportive care as compared to supportive care alone. Supportive care could include non-pharmacologic and pharmacologic, but not disease-modifying, interventions. We report results from two perspectives: a health care sector perspective (i.e., focusing on the direct medical care costs and health outcomes of the patient) and a modified societal perspective (i.e., including patient productivity impacts, caregiver time spent caregiving, caregiver quality of life, and caregiver direct medical costs). In alignment with our approach from our prior AD review, these perspectives are presented as co-base-case analyses given the enormity of the societal costs in AD. We adapted our AD decision analytic model developed for our prior AD assessment that included aducanumab.? The model consisted of five health states that tracked the severity of disease, including mild cognitive impairment (MCI) due to AD, mild AD, moderate AD, severe AD, and death. Figure 4.1. on the following page displays each of these health states and the possible transitions between them. Patients in the MCI due to AD health state could progress to the mild AD health state or stay in the MCl due to AD health state. Individuals in the mild AD health state could either stay in the mild AD health state, progress to either the moderate AD or severe AD health states or revert back to the MCl due to AD health state. Individuals in the moderate AD health state could either stay in the moderate AD health state, progress to the severe AD health state, or revert to the mild AD health state. Individuals in the severe AD health state could either stay in the severe AD health state or revert to the moderate AD health state. All health states could transition to the death health state due to all-cause or disease-specific mortality. Model cycle length was one year as has been used in prior published economic models and in clinical evidence.>*%? Specific to each health state, the model also tracked the setting of care (e.g., community or long-term care). Patients could transition from community to long-term care; however, once in long-term care, they remained there over the lifetime time horizon. The model was developed in Microsoft Excel. ©lnstitute for Clinical and Economic Review, 2023 Page 22 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Figure 4.1. Model Structure N //MCI dueto / \\_\ AD - / [ N e B S S o A f\f/ Mild AD -( Moi%rate \ SevereAD ) kS 7 S N D) AD: Alzheimer's disease, MCl: mild cognitive impairment The model included a hypothetical cohort of patients with MCI due to AD or mild AD entering the model and receiving either the intervention or comparator treatments. The hypothetical cohort had amyloid positivity confirmed by a reliable method prior to initiating treatment. Baseline population characteristics can be found in the Supplement. Consistent with population estimates, slightly more than half (55%) of the cohort started in the MCl due to AD health state, with the remaining cohort (45%) starting in the mild AD health state. Patients could progress to more severe AD health states over the model time horizon. The majority of the cohort (92%) started the model in a community setting of care. Model outcomes included quality-adjusted life years (QALYs), equal- value life years {evLYs), life years (LYs), years living outside of long-term care, and costs over a lifetime time horizon. All future costs and outcomes were discounted at 3% per year. Since the posting of our draft report, we have updated the baseline characteristics of the model to no longer include evidence from donanemab and we have updated the price for lecanemab now that the wholesale acquisition cost is available. No changes were made to the cost-effectiveness analysis since the evidence report. ©lnstitute for Clinical and Economic Review, 2023 Page 23 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 4.2. Key Model Assumptions and Inputs Table 4.1. reports model assumptions along with their rationale that are important to consider when interpreting the findings. A full list of model assumptions can be found in the Supplement. Table 4.1. Key Model Assumptions Assumption Rationale Lecanemab was effective at slowing the progression of disease while a patient had MCI due to AD or mild AD. Lecanemab was no longer effective once a patient reached moderate AD. The evidence on clinical outcomes that exists for lecanemab is in early AD. Based on trials of anti- amyloid therapies suggesting no benefit in moderate AD, clinical experts suggested there is likely no effect with anti-amyloid therapies at reducing disease progression once a patient has reached moderate AD. This assumption was tested in scenario analyses. Patients stopped receiving lecanemab treatment once they reached moderate AD. Based on trials of anti-amyloid therapies suggesting no benefit in moderate AD, clinical experts suggested there is likely no effect with these anti-amyloid treatments at reducing disease progression once a patient has reached moderate AD, and therefore in our model, treatment stopped once a patient reached moderate AD. In a scenario analysis, we modeled patients stopping treatment once they reached severe AD. Robust evidence is lacking on lecanemab's effect on clinical outcomes after a patient has stopped the treatment, and thus no additional clinical benefit was assumed after a patient stopped treatment. All occurrences of ARIA and its associated consequences (on cost, quality of life, and treatment discontinuation) were modeled in the first year of treatment. ARIA has been observed as an adverse event for many studied treatments that target aggregated beta- amyloid. Consistent findings across these studies suggest ARIA occurs early in the treatment course. AD: Alzheimer's disease, CDR-SB: Clinical Dementia Rating-Sum of Boxes, MCI: mild cognitive impairment Model inputs were identified from best-available evidence and stakeholder engagement. The primary clinical inputs included the transition probabilities among alive health states, mortality, progressions to long-term care, treatment efficacy, the occurrence of adverse events, and discontinuation. Utility estimates were retrieved for both the patient and caregiver. The primary cost inputs included intervention acquisition costs, administration costs, monitoring costs, adverse event costs, long-term care costs, and other patient medical and pharmacy costs. Costs to inform the societal perspective included patient productivity, caregiver productivity, and caregiver health care costs. Table 4.2. reports key model inputs, but an exhaustive description of all model inputs and their sources can be found in the Supplement. OInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page 24 Return to Table of Contents Table 4.2. Key Model Inputs Treatment Effectiveness* on Slowing Disease Progression: Lecanemab MCI due to AD 0.69 Mild AD 0.69 Moderate AD 1.00 Severe AD 1.00 ARIA: Lecanemab Probability of Any ARIA 21.5% Probability of Symptomatic ARIA 3.5% Probability of AE-Related Discontinuation 6.9% Patient Disutilities: Community Long-Term Care MCI due to AD -0.17 -0.17 Mild AD -0.22 -0.19 Moderate AD -0.36 -0.42 Severe AD -0.53 -0.59 Caregiver Disutilities: Community Long-Term Care MCI due to AD -0.03 -0.03 Mild AD -0.05 -0.05 Moderate AD -0.08 -0.08 Severe AD -0.10 -0.10 Cost Inputs: Lecanemab Annual Wholesale Acquisition Cost $26,500 MRI Unit Cost $261 Annual Cost of Long-term Care 588,728 AD: Alzheimer's disease, AE: adverse event, ARIA: amyloid-related imaging abnormalities, MCI: mild cognitive impairment, MRI: magnetic resonance imaging, PET: positron emission tomography *Only applied to health state progressions (i.e., transitions to more severe health states). 4.3. Results Base-Case Results Table 4.3. reports the model outcomes from the health care sector perspective and Table 4.4. reports the model outcomes from the societal perspective. Lecanemab was associated with more costs than the comparator, but also increases in life years, QALYs, evlLYs, and years in the community. Table 4.3. Model Outcomes for the Health Care Sector Perspective Intervention ; Years in the Treatment Cost* Total Cost | Life Years QALYs evLYs Community Lecanemab $109,000 $489,000 | 6.23 3.84 3.96 4.20 Supportive Care $0 $363,000 | 5.77 3.34 3.34 3.69 evlLYs: equal-value life years, QALYs: quality-adjusted life years *Intervention cost doesn't include provider administered mark-up (modeled as 6%), monitoring costs, or administration costs. OInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page 25 Return to Table of Contents The absolute QALYs and evlYs are lower in the societal perspective (Table 4.4.) due to the additional disutility added in this perspective to capture the reduction in quality of life for the caregiver. Table 4.4. Model Outcomes for the Modified Societal Perspective Intervention . Years in the Treatment Cost* Total Cost | Life Years QALYs evLYs Community Lecanemab $109,000 $790,000 | 6.23 3.49 3.64 4.20 Supportive Care S0 $670,000 | 5.77 2.98 2.98 3.69 evlLYs: equal-value life years, QALYs: quality-adjusted life years *Intervention cost doesn't include provider administered mark-up {(modeled as 6%), monitoring costs, or administration costs. Table 4.5. reports the incremental cost-effectiveness ratios for the base-case from the health care sector perspective and from the modified societal perspective. The modified societal perspective incremental cost-effectiveness ratios are slightly more favorable than those from the health care sector perspective. Table 4.5. Base-Case Incremental Cost-Effectiveness Ratios Lecanemab vs. Supportive Care Cost per Life Year Gained Cost per QALY Gained Health Care Sector $277,000 $254,000 Societal $265,000 $236,000 evlY: equal-value life year, QALY: quality-adjusted life year Perspective Cost per evLY Gained $204,000 $183,000 Sensitivity Analyses Figure 4.2. displays the tornado diagram from the one-way sensitivity analysis comparing lecanemab in addition to supportive care as compared to supportive care alone. The vertical axis is the incremental cost per QALY gained from the health care sector perspective. The blue coloring represents the range in incremental cost-effectiveness ratios from the base-case estimate to the incremental cost-effectiveness ratio corresponding to the lower bound of that model input. The green coloring represents the range in incremental cost-effectiveness ratios from the base-case estimate to the incremental cost-effectiveness ratio corresponding to the upper bound of that model input. As depicted in the figure, cost-effectiveness is primarily dependent on the effectiveness of the treatment in slowing progression of disease. Table 4.6. reports the lower and upper input, along with their associated incremental cost-effectiveness ratios, for the ten most influential model inputs from the one-way sensitivity analysis. OInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page 26 Return to Table of Contents Figure 4.2. One-Way Sensitivity Analysis Results: Lecanemab versus Supportive Care Alone, Health Care Sector Perspective S0 $250,000 $500,000 $750,000 $1,000,000 Hazard ratio on slowing progression of disease $162,225 -_ $960,615 Percent of MCl alive patients moving to Mild $241,816 II $§271,283 Patient disutility Severe AD, long-term care setting $239,968 II $269,328 Percent of Mild alive patients moving to Moderate $242,990 II $269,055 Patient disutility MCI, community care setting $245,572 II $262,603 Relative Risk of death from MCI $247,433 " $260,701 Patient disutility Mild AD, community care setting $247,968 " $259,917 Patient disutility Moderate AD, community care setting $248,053 " $259,824 Annual medical cost multiplier for MCI $248,410 " $259,742 Annual medical cost multiplier for Mild $249,211 || $258,860 mLow mHigh AD: Alzheimer's disease, MCIl: mild cognitive impairment ©Institute for Clinical and Economic Review, 2023 Page 27 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Table 4.6. Tornado Diagram Inputs: Lecanemab versus Supportive Care Alone, Health Care Sector Perspective Low Input Upper Input Lower Upper Model Input ICER ICER Input Input Hazard Ratio on Slowing Progression of Disease $162,000 $961,000 0.49 0.92 Percent of MCI Alive Patients Moving to Mild AD $271,000 $242,000 19% 28% Patient Disutility Severe AD, Long-Term Care Setting 5240,000 5269,000 -0.71 -0.47 Percent of Mild AD Alive Patients Moving to Moderate AD 5269,000 5243,000 28% 42% Patient Disutility MCI, Community Care Setting $263,000 $246,000 -0.20 -0.14 Relative Risk of Death from MCI $247,000 $261,000 1.48 2.19 Patient Disutility Mild AD, Community Care Setting 3260,000 5248,000 -0.26 -0.18 Patient Disutility Moderate AD, Community Care Setting 3248,000 260,000 -0.43 -0.29 Annual Medical Cost Multiplier for MCI $248,000 $260,000 1.00 1.35 Annual Medical Cost Multiplier for Mild AD $249,000 $259,000 1.27 1.88 AD: Alzheimer's disease, MCl: mild cognitive impairment *Standard error was estimated based on the 95% confidence interval for the primary endpoint of the pivotal trial. Table 4.7. reports results from the probabilistic sensitivity analysis for lecanemab versus supportive care from the health care sector perspective and the modified societal perspective. Other findings from the probabilistic sensitivity analyses, including credible intervals for lecanemab versus supportive care, can be found in the Supplement. Table 4.7. Probabilistic Sensitivity Analysis Results: Lecanemab versus Supportive Care Health Care Sector Cost Effective at Cost Effective at Cost Effective at Cost Effective at Perspective $50,000 $100,000 $150,000 $200,000 Per QALY Gained | 0% 0% 1% 21% Per evLY Gained | 0% 0% 11% 50% Madified Societal Cost Effective at Cost Effective at Cost Effective at Cost Effective at Perspective $50,000 $100,000 $150,000 $200,000 Per QALY Gained | 0% 0% 1% 30% Per evLY Gained | 0% 1% 25% 63% evlY: equal-value life year, QALY: quality-adjusted life year Scenario Analyses We conducted numerous scenario analyses to test the structural assumptions that were made. Details on the methods and results of each scenario analysis can be found in the Supplement. In Table 4.8, we highlight the findings from the scenario that assumed treatment stopped once a patient reached severe AD with the same assumed treatment effectiveness for moderate AD as OInstitute for Clinical and Economic Review, Alzheimer's Disease - Final Evidence Report 2023 Page 28 Return to Table of Contents what was modeled in MCl due to AD and mild AD. In this scenario, the incremental clinical outcomes are greater than they were in the base-case because the treatment benefit was assumed to persist through moderate AD, but the treatment costs and non-intervention costs are also greater in this scenario than they were in the base-case. Table 4.8. Incremental Cost-Effectiveness Ratios Assuming Treatment Continues until Severe AD Lecanemab vs. Supportive Care Perspective Cost per Life Year Gained Cost per QALY Gained Cost per evLY Gained Health Care Sector $293,000 $286,000 $226,000 Societal $290,000 $278,000 $211,000 evlLY: equal-value life year, QALY: quality-adjusted life year Threshold Analyses Table 4.9. reports the annual threshold prices for lecanemab from both the health care sector perspective and the modified societal perspective. Table 4.9. Threshold Analysis* Results for Lecanemab Health Care Sector Annual Price to Annual Price to Annual Price to Annual Price to Perspective Achieve $50,000 | Achieve $100,000 | Achieve $150,000 | Achieve $200,000 Threshold Threshold Threshold Threshold Per QALY Gained $3,200 $8,900 $14,600 $20,300 Per evlLY Gained $4,600 $11,700 $18,800 $26,000 Modified Societal Annual Price to Annual Price to Annual Price to Annual Price to Perspective Achieve $50,000 | Achieve $100,000 | Achieve $150,000 | Achieve $200,000 Threshold Threshold Threshold Threshold Per QALY Gained $4,600 $10,500 $16,400 $22,200 Per evlLY Gained $6,300 $13,900 $21,500 $29,100 evlY: equal-value life year, QALY: quality-adjusted life year *These threshold prices do not include the provider-administered markup, which was modeled as 6%. OInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page 29 Return to Table of Contents Model Validation We used several approaches to validate the model. First, we provided the preliminary model structure, methods and assumptions to manufacturers, patient groups, and clinical experts. Based on feedback from these groups, we refined data inputs used in the model, as needed. Second, we varied model input parameters to evaluate face validity of changes in results. We performed model verification for model calculations using internal reviewers. As part of ICER's efforts in acknowledging modeling transparency, we also shared the model with the relevant manufacturers for external verification shortly after publishing the draft Evidence Report. Finally, we compared results to other cost-effectiveness models in this clinical area. Uncertainty and Controversies There are important evidence uncertainties relevant to generating model outcomes, most of which relate to the effectiveness of the treatment on slowing progression of disease. For lecanemab, a hazard ratio on the progression to the next stage of dementia was available, but a confidence interval was not available to fully understand and capture the uncertainty in this estimate. In addition to uncertainty in the effect of each treatment on the progression of disease, there are other inputs in the model that have uncertainty. For example, the utilities for the patient and caregiver are from cross-sectional studies. Limitations of these studies include representing cross- sectional utility weights to estimate impacts of an individual's health state that may change over time and using instruments that might not be sensitive enough to detect AD-specific effects and/or second order effects for the caregivers. Additionally, the utility evidence is from a study that was published more than 20 years ago, so there is a potential that health systems and disease consequences have changed since then. We have conducted extensive sensitivity and scenario analyses, although there may be variation outside of what was modeled. We presented the modified societal perspective as a co-base-case analysis in this report due to the large impact of AD on caregivers, represented in the model by a disutility for caregivers and a large loss of caregiver productivity outside of the health care system. However, the cost effectiveness in the modified societal perspective did not greatly differ from analyses performed using the health care system perspective. This result may seem counterintuitive, but is largely the result of these treatments slowing the progression of AD, not stopping the progression or curing the AD. In addition, keeping a patient in earlier AD states longer, which delays the transition to long-term care, can increase productivity losses for the caregiver. These countervailing factors reduce the spread between the cost-effectiveness results using the health care system and modified societal perspectives. This highlights the complexities of capturing caregiver perspectives in the modified societal perspective in that caregivers may prefer to keep loved ones at home, rather than in a long- term care facility, although doing so may increase the negative financial impact on the caregiver. ©Institute for Clinical and Economic Review, 2023 Page 30 Alzheimer's Disease - Final Evidence Report Return to Table of Contents We do not include aducanumab as an intervention in this cost-effectiveness analysis given the lack of new clinical evidence and the numerous assumptions we included in our prior review. The assumptions we make in the base-case of this review align most closely with the optimistic treatment benefit scenario conducted in our prior AD review, but differences exist in model assumptions and programming between our prior review and this review. Finally, some commentators have suggested that thresholds should be adjusted for disease severity.®3 Their work suggests a threshold higher than $100,000 to $150,000 per QALY gained for severe conditions (like AD). However, thresholds much lower than $100,000 to $150,000 per QALY gained are suggested for less severe conditions. Specific methods by which to assign lower thresholds to some conditions and higher thresholds to others have not gained consensus in health economics, in part because they require a view of a single societal value for severity, and also because any divergence in thresholds creates "winners and losers," with equal health gains for some patients viewed as worth "less" than those of others. We present results at multiple cost- effectiveness thresholds but continue to provide an emphasis on results between $100,000- $150,000 per evLYG and QALY gained. 4.4 Summary and Comment At the current wholesale acquisition cost, lecanemab exceeds commonly cited thresholds. The cost-effectiveness findings are primarily driven by the effectiveness of lecanemab at slowing the progression of disease. ©Institute for Clinical and Economic Review, 2023 Page 31 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 5. Contextual Considerations and Potential Other Benefits Our reviews seek to provide information on potential other benefits offered by the intervention to the individual patient, caregivers, the delivery system, other patients, or the public that was not available in the evidence base nor could be adequately estimated within the cost-effectiveness model. These elements are listed in the table below, with related information gathered from patients and other stakeholders. Following the public deliberation on this report the appraisal committee will vote on the degree to which each of these factors should affect overall judgments of long-term value for money of the intervention in this review. Table 5.1. Contextual Considerations Contextual Consideration Relevant Information Acuity of need for treatment of individual patients based on short-term risk of death or progression to permanent disability The acuity of need is high, as AD is a progressive neurodegenerative disease that places a high burden on patients with the disease, caregivers, and society. Currently there is only one FDA-approved disease-modifying therapy. Magnitude of the lifetime impact on individual patients of the condition being treated AD has a moderate lifetime impact on patients with the disease. Delaying or stopping progression of AD would improve the quality and, potentially, the length of life of patients. However, late-onset AD affects individuals over the age of 65 and early-onset AD affects only a minority of individuals with the disease. Thus, unlike diseases that impact the patient's entire lifespan, AD has a large effect on a portion of an individual's lifespan, leading to our assessment of moderate lifetime impact. ©Institute for Clinical and Economic Review, 2023 Page 32 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Table 5.2. Potential Other Benefits or Disadvantages Potential Other Benefit or Disadvantage Relevant Information Patients' ability to achieve major life goals related to education, work, or family life AD has a substantial impact on patient independence for activities of daily living such as driving, shopping, financial tasks, etc. While most patients develop AD later in life after they have completed their education and may have left the workforce, delaying progression of the disease may have a significant impact on family life, particularly in patients with early-onset AD. Caregivers' quality of life and/or ability to achieve major life goals related to education, work, or family life Delaying progression of AD with anti-amyloid therapy could potentially decrease caregiver impact and stress, increasing caregiver ability to achieve major life goals. Caregivers tend to be younger than patients with AD, and thus the magnitude of benefit to caregivers may be larger over the lifetime than for patients. Patients' ability to manage and sustain treatment given the complexity of regimen Current anti-amyloid therapies require monthly or biweekly intravenous infusions, as well as MRIs for monitoring for side effects. This may place a significant burden on patients and caregivers, particularly for those who have difficulty accessing this level of care (e.g., living in a rural area, lack of transportation, caregivers who aren't able to take time off work, etc.), and as the disease progresses. Society's goal of reducing health inequities The impact of lecanemab on health inequities is unclear. Underrepresented minority groups such as Black and Hispanic populations have a higher prevalence of disease and are diagnosed at later stages, thus an effective treatment could potentially decrease disparities. Additionally, an effective disease-modifying drug could raise awareness of the disease and increase early-stage diagnosis of the disease. However, such groups are less well represented in clinical trials, and the drugs were not tested in people with moderate or severe AD, thus whether the drug has a differential impact in such populations is not known. In highlighting inequalities in the Alzheimer's disease space, ICER calculated the Health Improvement Distribution Index, looking at the relative proportion of any health gains from treatment of AD for the following groups who have a higher prevalence than the general US population. For more information on how we calculate the Health Improvement Distribution Index, refer to the Supplement. e African American = 1.6 ©Institute for Clinical and Economic Review, 2023 Page 33 Alzheimer's Disease - Final Evidence Report Return to Table of Contents CTAF VOTES At the public meeting, the CTAF deliberated and voted on the relevance of specific potential other benefits and contextual considerations on judgments of value for the interventions under review. The results of the voting are shown below. Further details on the intent of these votes to help provide a comprehensive view on long-term value for money are provided in the ICER Value Assessment Framework. When making judgments of overall long-term value for money, what is the relative priority that should be given to any effective treatment for early Alzheimer's disease with evidence of Alzheimer's disease pathology, on the basis of the following contextual considerations: Table 5.3. CTAF Votes on Contextual Considerations High priority Very high priority Low priority Very Low Priority Average Contextual Consideration . priority Acuity of need for treatment of individual patients based on short-term risk of death or 0 2 3 5 4 progression to permanent disability* Magnitude of the lifetime impact on individual patients of the condition being treated *One panel member did not vote Based on perspectives shared by the patient and clinical experts, a majority voted that given the acuity of need for treatment of individual patients with early Alzheimer's disease, high or very high priority should be given to any treatment. A majority of the panel also agreed on assigning very high priority for any effective treatment regarding the magnitude of the lifetime impact of early Alzheimer's disease. What are the relative effects of lecanemab added to supportive care versus supportive care alone on the following outcomes that inform judgment of the overall long-term value for money of lecanemab added to supportive care? Table 5.4. CTAF Votes on Potential Other Benefits or Disadvantages Questions Potential Other Benefit or Disadvantage Major Negative Effect Minor Negative Effect No Difference Minor Positive Effect Major Positive Effect Patients' ability to achieve major life goals related to education, work, or family life 10 Caregivers' quality of life and/or ability to achieve major life goals related to education, work, or family life Society's goal of reducing health inequities A majority of the panel voted that lecanemab added to supportive care would have a minor positive effect on patients' ability to achieve major life goals related to education, work, or family life. For ®lInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page 34 Return to Table of Contents this vote, the panel considered the importance of being able to live independently, able to travel, and having the opportunity to continue doing activities they are passionate about, as referenced by a patient expert at the public meeting. About half of the panel voted for a minor positive effect on caregivers' quality of life, considering stories shared by caregivers of needing to quit their job to take care of their loved one as well as impacts on their quality of life. While caregiving is often discussed in time or money spent, the panel also considered other aspects such as the loss of a patient's ability to communicate and how challenging it can be to understand the needs of the patient as the disease progresses. A majority of the panel agreed on lecanemab having a major negative effect towards society's goal of reducing health inequities. While panel members agreed that new treatments for early Alzheimer's disease have the opportunity to reduce inequities given the higher prevalence rates for Black and Hispanic individuals, the panel also raised concerns that there is already a lack of dementia care and they fear that the infrastructure needed to administer this drug could take investment away from other services. Panel members also acknowledged that higher-income patients may be favored for access to treatment given the added costs associated with this treatment, such as travel to infusion centers and brain imaging, and the burden this could place on Medicaid systems. Additionally Black and Hispanic patients are often diagnosed in later stages of dementia when it would be too late for this treatment, limiting access. They agreed that there should be efforts to reduce inequities. OlInstitute for Clinical and Economic Review, 2023 Page 35 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 6. Health Benefit Price Benchmarks Health Benefit Price Benchmarks (HBPBs) for the annual cost of treatment with lecanemab are presented in Table 6.1 below. The HBPB for a drug is defined as the price range that would achieve incremental cost-effectiveness ratios between $100,000 and $150,000 per QALY or per evLY gained. The health benefit price benchmark for lecanemab ranged from $8,900 to $21,500, or a 19% to 66% discount from lecanemab's wholesale acquisition cost. Table 6.1. Annual Cost-Effectiveness Threshold Prices* for Lecanemab Health Care . . Discount from WAC Sector Annual WAC Annual Price at Annual Price at to Reach Threshold . $100,000 Threshold | $150,000 Threshold . Perspective Prices QALYs Gained $26,500 $8,900 $14,600 45%-66% evLYs Gained $26,500 $11,700 $18,800 29%-56% Societal Annual WAC Annual Price at Annual Price at I:;s;:::'t' flr:rr:s'i'::: Perspective $100,000 Threshold | $150,000 Threshold Prices QALYs Gained $26,500 $10,500 $16,400 38%-60% evlLYs Gained $26,500 $13,900 $21,500 19%-48% WAC: wholesale acquisition cost, evLY: equal value life year, QALY: quality-adjusted life year *Threshold prices do not include any provider-administered mark-up, which was assumed to be 6% in the model. CTAF Votes Table 6.2. CTAF Votes on Long-Term Value for Money at Current Prices Question Low long-term value for money at current price Intermediate long-term value for money at current price High long- term value for money at current price Given the available evidence on comparative effectiveness and incremental cost-effectiveness, and considering other benefits, disadvantages, and contextual considerations, what is the long-term value | 15 0 0 for money of treatment at current pricing with lecanemab added to supportive care versus supportive care alone? The panel unanimously voted that lecanemab added to supportive care at its current price provides low long-term value for money. Panel members noted concerns about the uncertain durability of benefit of lecanemab treatment, given the lack of long-term data. Page 36 Return to Table of Contents ®Institute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report 7. Potential Budget Impact 7.1. Overview of Key Assumptions We used results from the cost-effectiveness model to estimate the potential total budgetary impact of lecanemab for people with MCI due to AD or mild AD. We used the wholesale acquisition cost and the three threshold prices from the health system perspective (at $50,000, $100,000, and $150,000 per QALY) in our estimates of budget impact. The aim of the potential budgetary impact analysis was to document the percentage of patients who could be treated at selected prices without crossing a potential budget impact threshold that is aligned with overall growth in the US economy. For 2022-2023, the five-year annualized potential budget impact threshold that should trigger policy actions to manage access and affordability is calculated to be approximately $777 million per year for new drugs. Potential budget impact was defined as the total differential cost of using each new therapy rather than relevant existing therapies for the treated population, calculated as differential health care costs (including drug costs) minus any offsets in these costs from averted health care events. All costs were undiscounted and estimated over a five-year time horizon. The budget impact analysis included the estimated number of patients with MCl due to AD or mild AD in the US who would be eligible for treatment with one of these interventions. Using our prior AD report and other analyses, we estimated there are 1.4 million patients in the US eligible for AD treatment that targets beta-amyloid. We assumed that 20% of these 1.4 million patients would initiate treatment in each of the five years, or approximately 280,000 people with AD per year. We assumed that lecanemab would be added on to standard of care for these individuals and that no current treatments would be displaced by use of the new treatments within the eligible population. 7.2. Results Figure 7.1. illustrates the health care sector perspective cumulative per-patient budget impact calculations for lecanemab compared to supportive care, based on the annualized wholesale acquisition cost of $26,500 per year of treatment. The average potential budgetary impact for lecanemab was approximately $27,000 per patient in year one, with the cumulative net cost increasing in years two through five as treatment continues, reaching approximately $104,000 by the end of the five-year horizon. The annual net cost slowly declined each year given various factors including treatment discontinuation and cost offsets. ©Institute for Clinical and Economic Review, 2023 Page 37 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Figure 7.1. Cumulative Per-Patient Budget Impact of Lecanemab $120,000 $100,000 $80,000 $60,000 $40,000 $20,000 I S0 1 2 3 4 5 Cost per Patient [taive Net Cumu Year Figure 7.2 illustrates the heath care sector perspective potential budget impact of lecanemab treatment, based on the annualized wholesale acquisition cost (526,500 per year of treatment), and the health care sector threshold prices to reach $50,000, $100,000, and $150,000 per QALY. Approximately 5% of the eligible population could be treated without crossing the ICER potential budget impact threshold of $777 million per year over five years at the annualized wholesale acquisition cost of $26,500 per year. Approximately 9% of patients could be treated without crossing the budget impact threshold at the $150,000 per QALY threshold price, increasing to approximately 16% of the population at the $100,000 per QALY threshold price, and increasing to approximately 59% of the population at the $50,000 per QALY threshold price. ©Institute for Clinical and Economic Review, 2023 Page 38 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Figure 7.2. Potential Budgetary Impact of Lecanemab Treatment $30,000 @ Base Case $25,000 &0 520,000 515,000 ® $150,000 per QALY Annual Treatment Price ® 550,000 per QALY 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percentage of Patients Treated Without Crossing Patential Budget Impact Threshaold Access and Affordability Alert Assuming lecanemab's current wholesale acquisition cost, approximately 5% of the roughly 1.4 million US patients eligible for AD treatment that targets beta-amyloid could be treated within five years without crossing the ICER potential budget impact threshold of $777 million per year. Therefore, at current pricing and projected uptake that is likely to exceed 5% of the eligible population, lecanemab's short-term potential budget impact exceeds our threshold. Additional efforts at achieving affordability and access must be considered, thus we are issuing an access and affordability alert. The purpose of an ICER access and affordability alert is to signal to stakeholders and policy makers that the amount of added health care costs associated with a new service may be difficult for the health system to absorb over the short term without displacing other needed services, creating pressure on payers to sharply restrict access, or causing rapid growth in health care insurance costs that would threaten sustainable access to high-value care for all patients. ©Institute for Clinical and Economic Review, 2023 Page 39 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 8. Policy Recommendations Following the CTAF's deliberation on the evidence, a policy roundtable discussion was moderated by ICER's president around how best to apply the evidence on the use of lecanemab for early Alzheimer's disease. The policy roundtable members included two patient advocates, two clinical experts, two payers, and one representatives from the drug maker. The discussion reflected multiple perspectives and opinions, and therefore, none of the statements below should be taken as a consensus view held by all participants. The top-line policy implications are presented below, and additional information can be found here. All Stakeholders Recommendation 1 All stakeholders have a responsibility and an important role to play in ensuring that new treatment options for patients with Alzheimer's disease are introduced in a way that addresses the impact on health inequities. AD is underdiagnosed and undertreated in the United States, with significant racial and ethnic disparities. Black and Hispanic Americans both have a higher risk of developing AD, are more likely to have delayed or missed diagnosis, and have more advanced cognitive and functional limitations at diagnosis than their non-Hispanic White counterparts.®* Additionally, individuals with limited English proficiency and persons with low education levels are also more likely to be underdiagnosed and diagnosed at later stages of cognitive dysfunction.®® As noted in the ICER report and presentation and confirmed by the vote of the CTAF panel, the evidence on balance of risks and benefits of lecanemab has significant limitations and may not be judged adequate for full regulatory approval. Even if the FDA does approve the drug under its traditional pathway as "safe and effective," CMS may judge that the evidence does not yet meet the standard for being "reasonable and necessary" and private insurers may judge that the evidence does not meet the standard for being "medically necessary." As described in a subsequent policy recommendation below, there are several options through which payers may choose to link coverage to requirements for further evidence development, which are likely to increase the risk that access to lecanemab will skew further toward patients and clinicians already connected with academic practice, resulting in poorer access for patients with lower economic resources, including many patients from Black and Hispanic communities. If lecanemab is covered without requirements for further evidence development, its introduction into practice will be prone to exacerbating health inequities due to several factors including the complexity in confirming eligibility for treatment, the lack of adequate tools for shared decision- ©Institute for Clinical and Economic Review, 2023 Page 40 Alzheimer's Disease - Final Evidence Report Return to Table of Contents making, and the potentially complicated logistics for receiving treatment and monitoring. Eligibility for treatment requires the diagnosis of MCl or mild AD, as well as confirmation of the presence of amyloid and evaluation for comorbidities that may heighten the risk of or be contraindications to treatment. Such evaluations are usually done by a dementia specialist, who is generally a neurologist or geriatrician. However, use of dementia specialists within one year of diagnosis is low, particularly amongst Black and Asian Americans 5, in part due to the shortage of such specialists.? Furthermore, payer coverage of the PET scans and ApoE genotype testing has been limited. Because lecanemab requires biweekly IV infusions, as well as MRIs to monitor for complications, timely access to both infusion centers and MRIs is essential. Studies have shown that barriers to timely medical care are often non-medical, such as lack of transportation, long waiting times, and lack of time off from work®":%8; these barriers can contribute to health inequities and increase caregiver impact. Finally, while CLARITY-AD included more patients from racial and ethnic minority communities than prior clinical trials testing anti-amyloid antibodies for AD, it was not representative of the US AD population and thus conclusions about the efficacy of lecanemab in subpopulations is limited. To address these concerns: Manufacturers should take the following actions: e Set initial prices according to value assessments from independent analysts to preserve access and affordability of new therapies. Fair pricing is required to fulfill the social responsibility held by manufacturers to avoid financial toxicity that falls hardest on the most underserved patients. Drug prices that are set well beyond the cost-effectiveness range can not only cause direct financial toxicity to patients, but also contribute to general health care cost growth that pushes families out of the insurance pool and causes rationing of care that may be harmful. Although Eisai should be praised for their transparency in presenting their model by which they calculated their launch price, studies have shown that industry- sponsored cost-effectiveness analyses are more likely to report favorable results for a new treatment.%® Thus, fair pricing should be based on independent cost-effectiveness analyses. e Work with communities and patient groups to develop reliable methods for recruiting diverse populations with MCI and AD for clinical trials and promote retention of such populations. The clinical trial population for CLARITY-AD included a more diverse patient population than prior studies of anti-amyloid therapies; however, patients from racial and ethnic minority communities were still underrepresented in the trial compared with the US population, as demonstrated by ICER's sample diversity rating of "fair". This limits the conclusions that can be drawn about the efficacy and safety of lecanemab in patients from such groups. Pavers should take the following actions: ©Institute for Clinical and Economic Review, 2023 Page 41 Alzheimer's Disease - Final Evidence Report Return to Table of Contents ¢ If considering coverage with evidence development, select options that balance the need for better evidence with the risks to equal access for patients with fewer economic resources and those from communities with less connection to academic clinicians and institutions. e Ensure that benefit designs developed in conjunction with employers and other plan sponsors do not create requirements for out-of-pocket spending that create major barriers to appropriate access for patients. For example, the out-of-pocket maximum for Part B services in Medicare is not capped, leading to a situation in which many patients will not be able to undertake certain treatments or will do so only with the guarantee of suffering significant financial hardship. Although many patients will carry supplemental insurance, close to six million Medicare beneficiaries do not, and millions more with Medicare Advantage have very high out-of-pocket maximums that they may not be able to afford. Oncology has been the primary example of this phenomenon, and it would be unconscionable should the advent of effective treatments for AD be accompanied by the extension of this same dysfunctional system. Lower out-of-pocket requirements obviously have broader financial repercussions on Medicare premiums and sustainability, and should be linked conceptually, and perhaps legislatively, with requirements for value-based pricing for infused agents. e Ensure coverage of PET scans and ApoE £4 genotype testing for the accurate diagnosis of AD and risk stratification for treatment with anti-amyloid antibodies, since only patients with demonstrated amyloid in the brain will be eligible for treatment and risk of ARIA may be greater in patients who are homozygous for the ApoE €4 mutation. e Recognize that because of a shortage of dementia specialists, in order to facilitate timely diagnosis and treatment of AD, continuation of the COVID pandemic-era expansion of telemedicine policies is necessary (e.g., allow for inter-state consultations, reimbursement for telehealth). However, due to the potential serious risks of treatment, expansion of tele- prescribing, rather than looser consultation-type arrangements should be the preferred option. e Consider wraparound programs that could help address barriers related to social determinants of health, such as transportation, case management, benefit counseling, legal assistance, education, and respite care. Use of wraparound programs has been associated with lower rates of hospitalization and emergency department use."° e Payers should follow the requirements set by the 2020 Improving HOPE for Alzheimer's Act® for CMS and provide outreach to providers and patients about the comprehensive care planning benefit for Medicare patients. ©Institute for Clinical and Economic Review, 2023 Page 42 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Health systems should take the following actions: Invest resources to increase capacity for screening and diagnosis. Improved access for screening and diagnosis across all segments of the patient population is an important goal to reduce existing disparities in dementia care. Actions to reduce disparities could include increasing access to dementia specialists in all communities through outreach clinics and telehealth; improving training, time support, and reimbursement for screening and diagnosis to be done in non-specialist settings (e.g., primary care); and supporting development of newer diagnostic testing such as blood-based biomarkers. Ensure that all interventions are appropriate for culturally and linguistically diverse populations and that interventions are accessible to low literacy populations. Such populations, due to social, economic, and cultural differences, may have different perceptions of illness and different goals of care."! Invest in infrastructure that will increase access to dementia care and infusion centers, particularly in rural or underserved areas. Most specialized dementia care occurs at large tertiary care centers. However, 85% of people were diagnosed with dementia by a non- specialist physician; less than one-quarter of those patients saw a dementia specialist within one year after diagnosis.®® Additionally, 20 states are considered "deserts" for neurologists, as a result, the average wait time to see a dementia specialist is 19 months.? Furthermore, dementia patients who live in rural areas receive less home health care and have shorter survival than their urban counterparts."? Thus, geographic expansion of dementia diagnosis and treatment services is essential to decreasing health inequities. This can be done through partnerships with local physicians and medical centers, as well as telehealth. Clinicians and clinical specialty societies should take the following actions: Given the shortage of dementia specialists to diagnose and manage patients with AD is likely only to grow, develop programs to educate and assist primary care providers in diagnosing and managing AD. Develop programs to recruit and retain a diverse workforce for the diagnosis and treatment of dementia. ©Ilnstitute for Clinical and Economic Review, 2023 Page 43 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Patient advocacy groups should take the following actions: e Develop programs to help deliver culturally sensitive information about AD diagnosis and treatment, and in collaboration with manufacturers and researchers, target the recruitment and retention of diverse populations for clinical trials. Multiple Stakeholders Recommendation 1 Patient groups, the manufacturer, clinicians, and clinical specialty societies should accurately describe the clinical benefits of lecanemab as a slowing of decline of cognition and function and avoid over-selling the potential benefit of treatment by using terms such as "improvement" or "return of quality of life" or "game changer" in all personal statements and advertising. Furthermore, there should not be an over-emphasis on the removal of amyloid from the brain, which still has not been conclusively linked to clinical outcomes. e Messaging from the manufacturer, clinical specialty societies, and patient groups, such as in patient-oriented websites and advertisements and provider education and detailing, should make it clear that lecanemab has not been shown to improve cognitive and functional performance. Rather, the messaging should state that lecanemab may slow the decline of cognition, function, and quality of life for patients and caregivers. e Clinicians and their patients should engage in shared decision-making founded upon a robust, individualized discussion of the potential harms and benefits of treatment. This should include discussion about the likelihood of benefit and risk of harm based on patients' individual clinical and social situation, uncertainty about whether removal of amyloid affects clinical outcomes, uncertainty about long-term harms, lack of benefit in moderate-to-severe AD, and potential financial toxicity. Many patients will have contraindications to therapy or a combination of comorbidities that should lead to very careful consideration of the risks and potential benefits for the individual. One common scenario is the active use of anticoagulant medication; while these patients were included in the trial, the safety and long-term outcomes of either stopping anticoagulation or taking lecanemab while anticoagulated must be weighed carefully for each individual patient. ©Institute for Clinical and Economic Review, 2023 Page 44 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Recommendation 2 Payers, healthcare systems, clinicians, and policymakers should work together to assure that the financial incentives associated with infusion delivery do not overly influence patient selection and prescribing. In order to ensure equitable access to infusions of lecanemab, there will be a need for more infusion centers, particularly in rural and underserved areas. There may be financial incentives for healthcare systems and clinicians to set up infusion centers. However, because of the potential risks of treatment, careful patient selection will be critical to maximize the benefit/harm ratio for the drug. Thus, stakeholders involved in delivery of infusions should build in safeguards to ensure that lecanemab is being prescribed only to patients who fit clinical criteria consistent with those receiving treatment benefit in the clinical trials, and only by prescribers with expertise in diagnosing and managing AD and potential treatment complications. Recommendation 3 All stakeholders would benefit from a robust yet practical evidence generation system linked to payer coverage in order to learn more about the real-world risks and benefits of lecanemab. CMS decided to employ coverage with evidence development {(CED) for anti-amyloid antibodies for AD under accelerated approval, and this decision currently extends to lecanemab while it is in the accelerated approval pathway. However, even if lecanemab is approved under a traditional FDA pathway, the balance of risks and potential benefits is narrow given questions about the longer- term durability of benefit and the real-world incidence and consequences of ARIA when the drug is used in broader populations. Whether or not the FDA requires a Risk Evaluation and Mitigation Strategy (REMS) to monitor safety, we recommend that stakeholders work together to establish a simple registry that will be able to provide rigorous data without creating access problems in diverse communities. There are a number of options that can be employed to generate additional post-approval real- world evidence. CED typically involves observational registries to collect standardized data on efficacy and safety. However, such registries are generally voluntary and collection of registry data can be time-consuming for providers, making it less likely that some providers will be willing to participate - potentially missing patients - and may require innovative approaches to funding the additional data collection (e.g., manufacturers, NIH or PCORI, clinical specialty societies). An example of an existing registry that could be leveraged is the Alzheimer's Network for Treatment & Diagnostics (ALZ-NET, https://www.alz-net.org). The US Department of Veterans Affairs (VA) has opted to require prescribers to participate in prospective medication use evaluation, also focused on safety. Finally, passive data collection using administrative claims or electronic health record data may also be employed; however, claims data often lack the clinical detail needed to fully ©Institute for Clinical and Economic Review, 2023 Page 45 Alzheimer's Disease - Final Evidence Report Return to Table of Contents assess efficacy and safety concerns. Although more patients may be captured through this strategy compared with voluntary registries, since the US does not have either a national all-payer claims database or a national electronic health record system, it would require agreement from all payers and/or health systems to share their data to a third party with analytic capabilities. Payers Recommendation 1 Although there is a tremendous need for disease-modifying treatment for AD, given that amyloid removal has not been conclusively linked to clinical benefit and that questions remain about the longer-term safety and effectiveness of lecanemab, it will be reasonable for payers to use prior authorization as a component of coverage of lecanemab, even if it receives full FDA approval. Prior authorization criteria for lecanemab should be based on clinical evidence and input from clinical experts and patient groups. The process for authorization should also be clear, accessible, efficient, and timely for providers. Manufacturers Recommendation 1 Manufacturers of future treatments for Alzheimer's disease should follow the example set by the manufacturer of lecanemab by sharing a transparent, explicit justification for their pricing. However, to foster affordability and good access for all patients, manufacturers should align prices with the patient-centered therapeutic value of their treatments as suggested in independent value assessments. Upon accelerated approval of lecanemab, when announcing the pricing of the drug, Eisai presented a cost-effectiveness model to justify the launch price of $26,500 for lecanemab. This kind of transparency regarding pricing should be standard when a new drug is launched or when the price of an existing drug is changed, and all manufacturers should follow the example of what Eisai did with explaining the price for lecanemab. However, the $26,500 price for lecanemab still falls above the value-based price range found in ICER's independent value assessment. This is likely in part due to the fact that industry-funded cost-effectiveness analyses are more likely to report favorable results for a new treatment.® Pricing beyond a drug's value threatens access for patients and causes financial toxicity. Aligning prices with independent assessments of the patient-centered value of treatments helps ensure that all patients who need treatment are able to access it at an affordable price. ©Ilnstitute for Clinical and Economic Review, 2023 Page 46 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Recommendation 2 Until longer-term data are available to address the significant uncertainties regarding the real- world risks and benefits of lecanemab, the manufacturer should provide all payers with a price similar to that negotiated with the VA that is at the lower range of value-based assessment. Drug prices that are set well beyond the cost-effective range cause not only financial toxicity for patients and families using the treatments, but also contribute to general health care cost growth that pushes families out of the insurance pool and causes others to ration their own care in ways that can be harmful. Manufacturers should therefore price novel treatments in accordance with the demonstrated benefits to patients. In settings of substantial uncertainty, initial pricing should err on the side of being more affordable. For example, the VA has negotiated a price for lecanemab that is below the list price and within the value-based price range suggested by the ICER's independent assessment. The manufacturer should provide other payers with the price negotiated by the VA, which would allow more patients access, generating additional data on the real-world effectiveness of novel treatments that could be used in future assessment updates. If longer-term data demonstrates a more substantial and more certain benefit/harm ratio, the manufacturer should be allowed to increase pricing in accordance with benefit. Recommendation 3 Manufacturers should release all patient-level data in order to help patients, clinicians, researchers, and regulators to understand more about the link between amyloid reduction and cognitive outcomes. The use of a surrogate outcome such as amyloid reduction is a common surrogate marker in clinical trials of anti-amyloid drugs, despite the fact that removal of amyloid has not been conclusively linked with clinical benefit in AD. Additionally, clinical trials typically report mean changes between groups. Aggregate measures can potentially obscure changes in individual patients that are important for assessing the magnitude of clinical benefit and for assessing the correlation between surrogate markers and clinical outcomes. Without patient-level data, clinicians have less ability to judge the relative harms and benefits for a particular patient (since aggregate results do not necessarily apply to individual patients), and patients are then forced to make decisions about treatment using data points that may or may not apply to them. For a drug with a small potential benefit and a benefit/harm ratio that depends highly on patient characteristics, this is the opposite of patient-centered care. Furthermore, without patient-level data, researchers and regulators are not only unable to perform independent analyses of the clinical trial results, but they are also not able to further explore the ©Institute for Clinical and Economic Review, 2023 Page 47 Alzheimer's Disease - Final Evidence Report Return to Table of Contents correlation between amyloid reduction and outcomes such as cognitive decline and quality of life. These kinds of analyses are crucial to advance the understanding of the pathophysiology of AD, and to understand whether amyloid reduction is an appropriate and useful surrogate outcome for future clinical trials. Thus, researchers and other stakeholders should push manufacturers for full patient-level data sets to be released, and the FDA should require such a data release to be standard for drug approvals. ©Institute for Clinical and Economic Review, 2023 Page 48 Alzheimer's Disease - Final Evidence Report Return to Table of Contents References 10. 11. 12, 13, 14, 15. 16. 17. 18. Rajmohan R, Reddy PH. Amyloid-Beta and Phosphorylated Tau Accumulations Cause Abnormalities at Synapses of Alzheimer's disease Neurons. J Alzheimers Dis. 2017;57(4):975- 999. 2023 Alzheimer's disease facts and figures. Alzheimers Dement. 2023. Deb A, Thornton JD, Sambamoorthi U, Innes K. Direct and indirect cost of managing alzheimer's disease and related dementias in the United States. Expert Rev Pharmacoecon Outcomes Res. 2017;17(2):189-202. UsAgainstAlzheimer's. AD PACE. usagainstalzheimers.org: UsAgainstAlzheimer's; 7/29/2020 2020. Petersen RC, Lopez O, Armstrong MJ, et al. Practice guideline update summary: Mild cognitive impairment. Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology. 2018;90(3):126-135. National Institute for Health and Care Excellence. Dementia: assessment, management, and support for people living with dementia and their carers - Interventions to promote cognition, independence, and wellbeing (NICE Guide NG97). 2019. Prillaman M. Alzheimer's drug slows mental decline in trial - but is it a breakthrough? Nature. 2022;610(7930):15-16. Lin GA WM, Synnott PG, McKenna A, Campbell J, Pearson, SD RD. Aducanumab for Alzheimer's Disease: Effectiveness and Value. Final Evidence Report and Meeting Summary. Institute for Clinical and Economic Review. 2021, Alzheimer's Association. 2022 Alzheimer's disease facts and figures. Alzheimers Dement. 2022;18(4):700-789. Mendez MF. Early-onset Alzheimer Disease and Its Variants. Continuum (Minneap Minn). 2019;25(1):34-51. Xing Y, Tang, Jia J. Sex Differences in Neuropsychiatric Symptoms of Alzheimer's Disease: The Modifying Effect of Apolipoprotein E epsilon4 Status. Behav Neurol. 2015;2015:275256. Kim S, Kim MJ, Kim S, et al. Gender differences in risk factors for transition from mild cognitive impairment to Alzheimer's disease: A CREDOS study. Compr Psychiatry. 2015;62:114-122. Mayeda ER, Glymour MM, Quesenberry CP, Whitmer RA. Inequalities in dementia incidence between six racial and ethnic groups over 14 years. Alzheimers Dement. 2016;12(3):216-224. Farrer LA, Cupples LA, Haines JL, et al. Effects of age, sex, and ethnicity on the association between apolipoprotein E genotype and Alzheimer disease. A meta-analysis. APOE and Alzheimer Disease Meta Analysis Consortium. JAMA. 1997;278(16):1349-1356. DeKosky ST, Scheff SW. Synapse loss in frontal cortex biopsies in Alzheimer's disease: correlation with cognitive severity. Ann Neurol. 1990;27(5):457-464. Bateman RJ, Xiong C, Benzinger TLS, et al. Clinical and biomarker changes in dominantly inherited Alzheimer's disease. N Engl J Med. 2012;367(9):795-804. Kim Y, Rosenberg P, Oh E. A Review of Diagnostic Impact of Amyloid Positron Emission Tomography Imaging in Clinical Practice. Dement Geriatr Cogn Disord. 2018;46(3-4):154-167. van Maurik IS, Broulikova HM, Mank A, et al. A more precise diagnosis by means of amyloid PET contributes to delayed institutionalization, lower mortality, and reduced care costs in a tertiary memory clinic setting. Alzheimers Dement. 2022. ©Institute for Clinical and Economic Review, 2023 Page 49 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. Rabinovici GD, Gatsonis C, Apgar C, et al. Association of Amyloid Positron Emission Tomography With Subsequent Change in Clinical Management Among Medicare Beneficiaries With Mild Cognitive Impairment or Dementia. JAMA. 2019;321(13):1286-1294. Boyle PA, Yu L, Schneider JA, Wilson RS, Bennett DA. Scam Awareness Related to Incident Alzheimer Dementia and Mild Cognitive Impairment: A Prospective Cohort Study. Ann intern Med. 2019;170(10):702-709. Schneider JA, Arvanitakis Z, Bang W, Bennett DA. Mixed brain pathologies account for most dementia cases in community-dwelling older persons. Neurology. 2007;69(24):2197-2204. Reuben DB, Tan ZS, Romero T, Wenger NS, Keeler E, Jennings LA. Patient and Caregiver Benefit From a Comprehensive Dementia Care Program: 1-Year Results From the UCLA Alzheimer's and Dementia Care Program. J Am Geriatr Soc. 2019;67(11):2267-2273. Gronek P, Balko S, Gronek J, et al. Physical Activity and Alzheimer's Disease: A Narrative Review. Aging Dis. 2019;10(6):1282-1292. DuZzLiY,LiJ, ZhouC, Li F, Yang X. Physical activity can improve cognition in patients with Alzheimer's disease: a systematic review and meta-analysis of randomized controlled trials. Clin Interv Aging. 2018;13:1593-1603. Gitlin LN, Kales HC, Lyketsos CG. Nonpharmacologic management of behavioral symptoms in dementia. JAMA. 2012;308(19):2020-2029. Kaduszkiewicz H, Zimmermann T, Beck-Bornholdt HP, van den Bussche H. Cholinesterase inhibitors for patients with Alzheimer's disease: systematic review of randomised clinical trials. BMYJ. 2005;331(7512):321-327. Matsunaga S, Kishi T, Iwata N. Combination therapy with cholinesterase inhibitors and memantine for Alzheimer's disease: a systematic review and meta-analysis. Int J Neuropsychopharmacol. 2014;18(5). Lin GA WM, Synnott PG, McKenna A, Campbell J, Pearson, SD RD. Aducanumab for Alzheimer's Disease: Effectiveness and Value; Evidence Report. Institute for Clinical and Economic Review. 2021. Sperling RA, Aisen, P. S., Beckett, L. A., Bennett, D. A., Craft, S., Fagan, A. M,, ... & Phelps, C. H. . Toward defining the preclinical stages of Alzheimer's disease: Recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. . Alzheimer's & dementia. 2011;7(3):280-292. Storandt M. Cognitive deficits in the early stages of Alzheimer's disease. Current Directions in Psychological Science. 2008;17(3):198-202. Howieson DB, Carlson, N. E., Moore, M. M., Wasserman, D., Abendroth, C. D., Payne-Murphy, J., & Kaye, J. A. . Trajectory of mild cognitive impairment onset. Journal of the International Neuropsychological Society. 2008;14(2):192-198. Lennon JC, Aita SL, Bene VAD, et al. Black and White individuals differ in dementia prevalence, risk factors, and symptomatic presentation. Alzheimers Dement. 2022;18(8):1461-1471. AARP, National Alliance for Caregiving. Caregiving in the United States 2020. Washington, D.C. May 2020 2021. Bose M, Tokarewich L, Bratches RWR, Barr PJ. Caregiving in a Diverse America: Beginning to Understand the Systematic Challenges Facing Family Caregivers. Washington, D.C.: National Alliance for Caregiving; November 2021 2021. Improving HOPE for Alzheimer's Act. In. Tonko P, Walorski J, Blumenauer E, Guthrie B, Waters M, Smith C, trans. 116th ed2020. van Dyck CH, Swanson CJ, Aisen P, et al. Lecanemab in Early Alzheimer's Disease. New England Journal of Medicine. 2022. ©Institute for Clinical and Economic Review, 2023 Page 50 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 37. 38. 39. 40. 41. 42. 43, 44, 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. Andrews IS, Desai U, Kirson NY, Zichlin ML, Ball DE, Matthews BR. Disease severity and minimal clinically important differences in clinical outcome assessments for Alzheimer's disease clinical trials. Alzheimers Dement (N Y). 2019;5:354-363. Lansdall C) MF, Butler LM, Delmar P, Pross N, Qin S, McLeod L, Zhou X, Kerchner GA, Doody RS. Establishing clinically meaningful change on outcome assessments frequently used in trials of mild cognitive impairment due to Alzheimer's disease. The Journal of Prevention of Alzheimer's Disease. 2022. Schrag A, Schott JM, Alzheimer's Disease Neuroimaging |. What is the clinically relevant change on the ADAS-Cog? J Neurol Neurosurg Psychiatry. 2012;83(2):171-173. Rockwood K, Fay S, M G. The ADAS-cog and clinically meaningful change in the VISTA clinical trial of galantamine for Alzheimer's disease. International Journal of Geriatric Psychiatry: A journal of the psychiatry of late life and allied sciences. 2010;25(2):191-201. Liu KY, Schneider LS, R. H. The need to show minimum clinically important differences in Alzheimer's disease trials. The Lancet Psychiatry. 2021;8(11):1013-1016. Cohen S. Context of Clarity AD Results. Paper presented at: Clinical Trials on Alzheimer's Disease2022; San Francisco. Bateman R. Imaging, Plasma, and CSF Biomarkers Assessments from Clarity AD. Paper presented at: Clinical Trials on Alzheimer's Diease2022; San Francisco. Sabbagh M. Safety Profile of Lecanemab in Early Alzheimers Disease. Paper presented at: Clinical Trials on Alzheimer's Disease2022; San Francisco. Sabbagh M, CH. vD. Response to: Multiple Cerebral Hemorrhages in a Patient Receiving Lecanemab and Treated with t-PA for Stroke. New England Journal of Medicine. 2023. Reish NJ JP, Stamm B, Flanagan ME, Sugg E, Tang M, Donohue KL, McCord M, Krumpelman C, Mesulam MM, Castellani R. . Multiple cerebral hemorrhages in a patient receiving lecanemab and treated with t-PA for stroke. The New England journal of medicine. 2023. Science. Scientists tie third clinical trial death to experimental Alzheimer's drug. Science Insider Web site. https://www.science.org/content/article/scientists-tie-third-clinical-trial-death- experimental-alzheimer-s-drug. Published 2022. Accessed February 2nd, 2023. van Dyck C. Lecanemab for the Treatment of Early Alzheimers Disease: Topline Efficacy Results from Clarity AD. Paper presented at: Clinical Trials on Alzheimer's Disease2022; San Francisco. Lacorte E, Ancidoni A, Zaccaria V, et al. Safety and efficacy of monoclonal antibodies for Alzheimer's disease: a systematic review and meta-analysis of published and unpublished clinical trials. Journal of Alzheimer's disease. 2022;87(1):101-129. Pang M, Zhu L, Gabelle A, et al. Effect of reduction in brain amyloid levels on change in cognitive and functional decline in randomized clinical trials: An instrumental variable meta-analysis. Alzheimer's & dementia. 2022. Villain N, Planche V, Levy R. High-clearance anti-amyloid immunotherapies in Alzheimer's disease. Part 1: Meta-analysis and review of efficacy and safety data, and medico-economical aspects. Rev Neurol (Paris). 2022. Ackley S, Glymour MM. Comment on "Effect of reduction in brain amyloid levels on change in cognitive and functional decline in randomized clinical trials: An updated instrumental variable meta-analysis". Alzheimers & Dementia. 2022:1-2. Alexander GC, Karlawish J. The Problem of Aducanumab for the Treatment of Alzheimer Disease. Ann Intern Med. 2021;174(9):1303-1304. Administration USFD. Incorporating Clinical Outcome Assessments into Endpoints for Regulatory Decision-Making. In. Patient-Focused Drug Development Guidance Workshop2022. ©Ilnstitute for Clinical and Economic Review, 2023 Page 51 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. Hauber B PR, Krasa HB, Vradenburg G, Comer M, Callahan LF, Winfield J, Potashman M, Hartry A, Lee D, Wilson H. Assessing What Matters to People Affected by Alzheimer's Disease: A Quantitative Analysis. Neurology and Therapy. 2023:1-23. Abdelnour C, Agosta F, Bozzali M, et al. Perspectives and challenges in patient stratification in Alzheimer's disease. Alzheimers Res Ther. 2022;14(1):112. Fleming TR, DeMets DL. Surrogate end points in clinical trials: are we being misled? Ann Intern Med. 1996;125(7):605-613. Rogers SL, Farlow MR, Doody RS, Mohs R, Friedhoff LT. A 24-week, double-blind, placebo- controlled trial of donepezil in patients with Alzheimer's disease. Donepezil Study Group. Neurology. 1998;50(1):136-145. Caro JJ, Getsios D, Migliaccio-Walle K, Raggio G, Ward A. Assessment of health economics in Alzheimer's disease (AHEAD) based on need for full-time care. Neurology. 2001;57(6):964-971. Green C, Handels R, Gustavsson A, et al. Assessing cost-effectiveness of early intervention in Alzheimer's disease: An open-source modeling framework. Alzheimers Dement. 2019;15(10):1309-1321. Neumann PJ, Hermann RC, Kuntz KM, et al. Cost-effectiveness of donepezil in the treatment of mild or moderate Alzheimer's disease. Neurology. 1999;52(6):1138-1145. Nguyen KH, Comans TA, Green C. Where are we at with model-based economic evaluations of interventions for dementia? a systematic review and quality assessment. International psychogeriatrics. 2018;30(11):1593-1605. Lakdawalla DN, Phelps CE. Health Technology Assessment With Diminishing Returns to Health: The Generalized Risk-Adjusted Cost-Effectiveness (GRACE) Approach. Value in Health. 2021;24(2):244-249. Lin PJ, Daly AT, Olchanski N, et al. Dementia Diagnosis Disparities by Race and Ethnicity. Med Care. 2021;59(8):679-686. Cooper C, Tandy AR, Balamurali TB, Livingston G. A systematic review and meta-analysis of ethnic differences in use of dementia treatment, care, and research. Am J Geriatr Psychiatry. 2010;18(3):193-203. Drabo EF, Barthold D, Joyce G, Ferido P, Chang Chui H, Zissimopoulos J. Longitudinal analysis of dementia diagnosis and specialty care among racially diverse Medicare beneficiaries. Alzheimers Dement. 2019;15(11):1402-1411. Butkus R, Rapp K, Cooney TG, Engel LS, Health, Public Policy Committee of the American College of P. Envisioning a Better U.S. Health Care System for All: Reducing Barriers to Care and Addressing Social Determinants of Health. Ann Intern Med. 2020;172(2 Suppl):S50-S59. Caraballo C, Ndumele CD, Roy B, et al. Trends in Racial and Ethnic Disparities in Barriers to Timely Medical Care Amaong Adults in the US, 1999 to 2018. JAMA Health Forum. 2022;3(10):e223856. Zhou T, Xie F. Sponsorship bias in oncology cost effectiveness analysis. J Clin Epidemiol. 2023;156:22-29. Vest JR, Harris LE, Haut DP, Halverson PK, Menachemi N. Indianapolis Provider's Use Of Wraparound Services Associated With Reduced Hospitalizations And Emergency Department Visits. Health Aff (Millwood). 2018;37(10):1555-1561. Chiong W, Tsou AY, Simmons Z, et al. Ethical Considerations in Dementia Diagnosis and Care. AAN Position Statement. 2021;97(2):80-89. Rahman M, White EM, Thomas KS, Jutkowitz E. Assessment of Rural-Urban Differences in Health Care Use and Survival Among Medicare Beneficiaries With Alzheimer Disease and Related Dementia. JAMA Netw Open. 2020;3(10):e2022111. ©Institute for Clinical and Economic Review, 2023 Page 52 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. McKhann GM, Knopman DS, Chertkow H, et al. The diagnhosis of dementia due to Alzheimer's disease: Recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimer's & Dementia. 2011;7(3):263-269. Albert MS, DeKosky ST, Dickson D, et al. The diagnosis of mild cognitive impairment due to Alzheimer's disease: Recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimer's & Dementia. 2011;7(3):270-279. Food and Drug Administration. Combined FDA and Applicant PCNS Drugs Advisory Committee Briefing Document. 2020. National Organization for Rare Disorders l. Superficial Siderosis. Rare Disease Database Web site. Published 2020. Accessed December 12, 2022. Gharbi-Meliani A, Dugravot A, Sabia S, et al. The association of APOE €4 with cognitive function over the adult life course and incidence of dementia: 20 years follow-up of the Whitehall Il study. Alzheimer's research & therapy. 2021;13(1):1-1. Cummings JL. Defining and labeling disease-modifying treatments for Alzheimer's disease. Alzheimers Dement. 2009;5(5):406-418. Chen GF, Xu TH, Yan Y, et al. Amyloid beta: structure, biology and structure-based therapeutic development. Acta Pharmacologica Sinica. 2017;38(9):1205-1235. Medeiros R, Baglietto-Vargas D, FM. L. The role of tau in Alzheimer's disease and related disorders. CNS neuroscience & therapeutics. 2011;17(5):514-524. Tanner JA, GD. R. Relationship between tau and cognition in the evolution of Alzheimer's disease: New insights from tau PET. Journal of Nuclear Medicine. 2021;62(5):612-613. Jack Jr CR, Knopman DS, Jagust W/, et al. Tracking pathophysiological processes in Alzheimer's disease: an updated hypothetical model of dynamic biomarkers. The lancet neurology. 2013;12(2). Drummond E, Pires G, MacMurray C, et al. Phosphorylated tau interactome in the human Alzheimer's disease brain. Brain. 2020;143(9):2803-2817. Pedrosa H, De Sa A, Guerreiro M, et al. Functional evaluation distinguishes MCI patients from healthy elderly people-the ADCS/MCI/ADL scale. The journal of nutrition, health & aging. 2010;14(8):703-709. Wang J, Logovinsky V, Hendrix SB, et al. ADCOMS: a composite clinical outcome for prodromal Alzheimer's disease trials. Journal of Neurology, Neurosurgery & Psychiatry. 2016;87(9):993-999. The EuroQol Group. EuroQol-a new facility for the measurement of health-related quality of life. Health Policy. 1990;16(3):199-208. Logsdon RG, Gibbons LE, McCurry SM, Teri L. Quality of Life-Alzheimer's Disease Scale (QOL-AD). In: PsycTests A, ed1999. Zarit SH, Reever KE, Back-Peterson J. Relatives of the impaired elderly: correlates of feelings of burden. The Gerontologist. 1980;20:649-655. Rajan KB, Weuve J, Barnes LL, McAninch EA, Wilson RS, Evans DA. Population estimate of people with clinical Alzheimer's disease and mild cognitive impairment in the United States (2020- 2060). Alzheimers Dement. 2021;17{12):1966-1975. DiBenedetti DB, Slota C, Wronski SL, et al. Assessing what matters most to patients with or at risk for Alzheimer's and care partners: a qualitative study evaluating symptoms, impacts, and outcomes. Alzheimers Res Ther. 2020;12(1):90. ©Institute for Clinical and Economic Review, 2023 Page 53 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. Reus VI, Fochtmann LJ, Eyler AE, et al. The American Psychiatric Association practice guideline on the use of antipsychotics to treat agitation or psychosis in patients with dementia. American Journal of Psychiatry. 2016;173(5):543-546. Jack CR, Ir.,, Bennett DA, Blennow K, et al. NIA-AA Research Framework: Toward a biological definition of Alzheimer's disease. Alzheimers Dement. 2018;14(4):535-562. Cummings J, Aisen P, Apostolova LG, Atri A, Salloway S, Weiner M. Aducanumab: Appropriate Use Recommendations. The Journal of Prevention of Alzheimer's Disease. 2021;8(4):398-410. Food and Drug Administration. Early Alzheimer's Disease: Developing Drugs for Treatment Guidance for Industry. https://www.fda.gov/media/110903/download. Published 2018. Accessed November 13, 2020. Cook DJ, Mulrow CD, Haynes RB. Systematic reviews: synthesis of best evidence for clinical decisions. Ann Intern Med. 1997;126(5):376-380. Higgins J, Thomas, J, Chandler, J, Cumpston, M, Li, T, Page, MJ, Welch, VA. Cochrane Handbook for Systematic Reviews of Interventions version 6.1 (updated September 2020). https://training.cochrane.org/handbook/current. Published 2020. Accessed. Moher D, Liberati A, Tetzlaff J, Altman DG, The Prisma Group. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLOS Medicine. 2009;6(7):e1000097. Sterne JAC, Savovic J, Page MJ, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;366. Higgins J, Savovi¢ J, Page M, Sterne J. Revised Cochrane risk-of-bias tool for randomized trials (RoB 2). 2019. Ollendorf DA, Pearson SD. An integrated evidence rating to frame comparative effectiveness assessments for decision makers. Med Care. 2010;48(6 Suppl}:5145-152. Ollendorf D, Pearson, SD. ICER Evidence Rating Matrix: A User's Guide. https://icer- review.org/methodology/icers-methods/icer-evidence-ratingmatrix/. . Published 2020. Updated January 31, 2020. Accessed. Swanson CJ, Zhang Y, Dhadda S, et al. A randomized, double-blind, phase 2b proof-of-concept clinical trial in early Alzheimer's disease with lecanemab, an anti-AB protofibril antibody. Alzheimer's research & therapy. 2021;13(1):1-4. Satlin A, Wang J, Logovinsky V, et al. Design of a Bayesian adaptive phase 2 proof-of-concept trial for BAN2401, a putative disease-modifying monoclonal antibody for the treatment of Alzheimer's disease. Alzheimer's & Dementia: Translational Research & Clinical Interventions. 2016;2(1):1-12. Dhadda S, Kanekiyo M, Li D, et al. Consistency of efficacy results across various clinical measures and statistical methods in the lecanemab phase 2 trial of early Alzheimer's disease. Alzheimer's Research & Therapy. 2022;14(1):1-9. Yagi T, Swanson C, Nakagawa M, et al. The Phase 2 study of BAN2401 ( anti-amyloid beta protofibril antibody) in early Alzheimer's disease. Paper presented at: 61st Annual Meeting of the Japanese Society of Neurology2020. Swanson CJ, Dhadda S, Irizarry M, et al. Clinical, Biomarker, and Safety Update from the Lecanemab Phase 2 Study. Lecanemab: An Assessment of the Clinical Effects, the Correlation of Plasma AB42/40 Ratio With Changes in Brain Amyloid PET SUVr, and Safety from the Core and Open Label Extension of the Phase 2 Proof-of-Concept Study, BAN2401-G000-201, in Subjects With Early Alzheimer's Disease. Paper presented at: CTAD Annual Meeting2021; Boston, MA. USA. ©Institute for Clinical and Economic Review, 2023 Page 54 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. Swanson CJ, Zhang Y, Dhadda S, et al. Treatment of early AD subjects with BAN2401, an anti-Ab protofibril monoclonal antibody, significantly clears amyloid plaque and reduces clinical decline. Alzheimers & Dementia. 2018;14(7):1668. Molinuevo JL, Koyama A, Dhadda S, et al. BAN2401 in early Alzheimer's disease: Neurodegeneration biomarker analysis from a randomized Phase 2 study. Paper presented at: Alzheimer's Association International Conference2019; Los Angeles, CA. Landry |, Kanekiyo M, Aluri J, et al. Lecanemab (BAN2401) infusion reactions and immunogenicity: results from randomized Phase 2 study and an open-label extension. Paper presented at: Alzheimer's Association International Conference (AAIC) Annual Meeting.2022; San Diego, CA. McDade E, Cummings JL, Dhadda S, et al. Lecanemab in patients with early Alzheimer's disease: detailed results on biomarker, cognitive, and clinical effects from the randomized and open- label extension of the phase 2 proof-of-concept study. Alzheimer's Research & Therapy. 2022;14(1):1-7. Swanson C, Dhadda S, Hodgkinson M, et al. Preliminary Analysis Of BAN2401 Effects On Brain Amyloid And ARIA-E Findings Over 12 Months Of Treatment In The Open-Label Extension Of The Phase2b Study BAN2401-G000-201 In Subjects With Early Alzheimer's Disease. Neurology. 2021;96. Rawal S, Doung A, Landry |, et al. Absolute Bioavailability of aSingle, Fixed Subcutaneous Dose ofLecanemab in Healthy Subjects. Paper presented at: Alzheimer's Association International Conference2022; San Diego, CA. Irizarry M. Clarity AD: Clinical Trial Background and Study Design Overview. Paper presented at: Clinical Trials on Alzheimer's Disease2022; San Francisco. DA B. Consistency of Efficacy Assessments Across Various Statistical Methods from the Lecanemab Phase 2 Proof-of-Concept Study, BAN2401-G000-201, in Subjects with Early Alzheimer's Disease. CTAD; 2022. Swanson CJ, Zhang Y, Dhadda S, et al. Persistence Of BAN2401-Mediated Amyloid Reductions Post-treatment: A Preliminary Comparison of Amyloid Status Between the Core Phase of BAN2401-G000-201 and Baseline of the Open-Label Extension Phase in Subjects with Early Alzheimer's Disease (1330). Neurology. 2020;94(15 Supplement):1330. M S. Safety Profile of Lecanemab in Early Alzheimers Disease. CTAD; 2022. Piller C. Scientists tie third clinical trial death to experimental Alzheimer's drug. Science Insider. 2022. Reyderman L, Hayato S, Reddy H, et al. Modeled Impact of APOE4 Genotype on ARIA-E Incidence in Patients Treated With Lecanemab. Alzheimer's & Dementia. 2022;18(510):e069402. Ackley SF, Zimmerman SC, Brenowitz WD, et al. Effect of reductions in amyloid levels on cognitive change in randomized trials: instrumental variable meta-analysis. BMJ. 2021;372:n156. NV, Planche V. Meta-Analysis of High Clearance Anti-Amyloid Immunotherapies Trials in Early Alzheimer's Disease: a Significant CLinical Effect but a High Risk/Benefit Ratio. CTAD; 2022. Sanders GD, Neumann PJ, Basu A, et al. Recommendations for Conduct, Methodological Practices, and Reporting of Cost-effectiveness Analyses: Second Panel on Cost-Effectiveness in Health and Medicine. Jama. 2016;316(10):1093-1103. Potashman M, Benea L, Gillis C, Gianinazzi M, lkram MA, Maserejian N. Estimating the prevalence of AB confirmed Alzheimer's disease using a funnel based approach. International Society of Pharmacoeconomics and Qutcomes Research Europe 2020; 2020; Virtual. ©Institute for Clinical and Economic Review, 2023 Page 55 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. Johnson RW. What is the lifetime risk of needing and receiving long-term services and supports? https://aspe.hhs.gov/basic-report/what-lifetime-risk-needing-and-receiving-long-term-services- and-supports. Published 2019. Accessed January 4, 2021. Potashman M, Buessing M, Levitchi Benea M, et al. Estimating progression rates across the spectrum of Alzheimer's disease for amyloid positive individuals using National Alzheimer's Coordinating Center data. CTAD 2020; 2020. Andersen K, Lolk A, Martinussen T, Kragh-Sgrensen P. Very mild to severe dementia and mortality: A 14-year follow-up - The Odense study. Dement Geriatr Cogn Disord. 2010;29(1):61- 67. Neumann PJ, Kuntz KM, Leon J, et al. Health utilities in Alzheimer's disease: a cross-sectional study of patients and caregivers. Med Care. 1999;37(1):27-32. Landeiro F, Mughal S, Walsh K, et al. Health-related quality of life in people with predementia Alzheimer's disease, mild cognitive impairment or dementia measured with preference-based instruments: a systematic literature review. Alzheimers Res Ther. 2020;12(1):154. Mintun MA, Lo AC, Duggan Evans C, et al. Donanemab in early Alzheimer's disease. New England Journal of Medicine. 2021;384(18):1691-1704. Xu R, Insinga RP, Golden W, Hu XH. EuroQol (EQ-5D) health utility scores for patients with migraine. Quality of Life Research. 2011;20(4):601-608. Mesterton J, Wimo A, By A, Langworth S, Winblad B, Jonsson L. Cross sectional observational study on the societal costs of Alzheimer's disease. Curr Alzheimer Res. 2010;7(4):358-367. IBM Micromedex. Redbook. www.micromedexsolutions.com. Published 2021. Accessed. Centers for Medicare & Medicaid Services. Physician Fee Schedule. 2022. Leibson CL, Long KH, Ransom JE, et al. Direct medical costs and source of cost differences across the spectrum of cognitive decline: a population-based study. Alzheimers Dement. 2015;11(8):917-932. Barthold D, Joyce G, Ferido P, et al. Pharmaceutical Treatment for Alzheimer's Disease and Related Dementias: Utilization and Disparities. J Alzheimers Dis. 2020;76(2):579-589. Administration on Aging. Costs of Care. https://longtermcare.acl.gov/costs-how-to-pay/costs-of- care.html#:~:text=Below%20are%20some%20national%20average,room%20in%20a%20nursing %20home. Published 2020. Accessed December 23, 2020. Bureau of Economic Analysis. Price Indexes for Personal Consumption Expenditures by Major Type of Product. In:2020. Robinson RL, Rentz DM, Andrews JS, et al. Costs of Early Stage Alzheimer's Disease in the United States: Cross-Sectional Analysis of a Prospective Cohort Study (GERAS-US)1. J Alzheimers Dis. 2020;75(2):437-450. Average hourly and weekly earnings of all employees on private nonfarm payrolls by industry sector, seasonally adjusted. 2020. Haro JM, Kahle-Wrobleski K, Bruno G, et al. Analysis of burden in caregivers of people with Alzheimer's disease using self-report and supervision hours. J Nutr Health Aging. 2014;18(7):677-684. Hauber B, DiBenedetti, D. A Quantitative Approach to Understanding What Matters Most to People With and at Risk for Alzheimer's Disease and to Care Partners. RTI Health Solutions. 2020. Pickard AS, Law EH, Jiang R, et al. United States Valuation of EQ-5D-5L Health States Using an International Protocol. Value in health : the journal of the International Society for Pharmacoeconomics and Outcomes Research. 2019;22(8):931-941. ©Institute for Clinical and Economic Review, 2023 Page 56 Alzheimer's Disease - Final Evidence Report Return to Table of Contents 142. Belger M, Dell'Agnello G, Enstone A, Wyn R, Tockhorn-Heidenreich A. PCR24 The Impact of Informal Caregiving in Alzheimer&#x2019;s Disease Dementia: A Health Utility Study in the UK. Value in Health. 2022;25(7):5545. 143. Tahami Monfared AA, Tafazzoli A, Chavan A, Ye W, Zhang Q. The Potential Economic Value of Lecanemab in Patients with Early Alzheimer's Disease Using Simulation Modeling. Neurol Ther. 2022;11(3):1285-1307. 144. Boustani M, Doty EG, Garrison LP, Jr., et al. Assessing the Cost-effectiveness of a Hypothetical Disease-Modifying Therapy With Limited Duration for the Treatment of Early Symptomatic Alzheimer Disease. Clin Ther. 2022. 145. Pearson SD, Towse A, Lowe M, Segel CS, Henshall C. Cornerstones of 'fair'drug coverage: appropriate cost sharing and utilization management policies for pharmaceuticals. Journal of Comparative Effectiveness Research. 2021;10(7):537-547. ©Ilnstitute for Clinical and Economic Review, 2023 Page 57 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Supplemental Materials ©Institute for Clinical and Economic Review, 2023 Page 58 Alzheimer's Disease - Final Evidence Report Return to Table of Contents A. Background: Supplemental Information Al. Definitions General definitions Alzheimer's disease {AD): A neurodegenerative brain disease with presenting symptoms including impairment in cognition (memory, language, executive function {e.g., problem-solving and completing tasks), and visuospatial function, all of which result in the loss of ability to perform activities of daily living (e.g., paying bills, bathing, dressing, etc.)."'* Changes in mood and personality, along with decreased or poor judgment and sleep disturbances, also occur. The main pathologies of AD are the accumulation of two abnormal protein deposits: protein tau tangles inside neurons and beta-amyloid plaques outside of the neurons in the brain. Stages of AD include a preclinical phase, MCI due to AD, and dementia due to AD.® Amyloid-related imaging abnormalities (ARIA): Abnormalities on MRI that can present as either edema/effusion (ARIA-E) or hemorrhage or superficial siderosis (ARIA-H), and is commonly associated with anti-amyloid therapies. ARIA may be asymptomatic {observed only on neuroimaging) or can cause symptoms such as headache, nausea, confusion, and gait abnormalities. ARIA is most commonly seen early on in a treatment period and is more frequently observed in ApoE €4 carriers as compared to non-carriers. Management of ARIA may include MRI monitoring, dose suspension or termination, treatment titration, etc."> o Hemorrhage: Bleeding from a damaged blood vessel. e Superficial siderosis: A degenerative disorder affecting the brain and spinal cord. Itis the result of persistent long-term bleeding into the subarachnoid space in the brain resulting in a build-up of hemosiderin on the brain surface and pia matter from cerebrospinal fluid."® Symptoms include hearing loss, movement abnormalities, and motor difficulties. Apolipoprotein g4 (ApoE £4): An allele that increases the risk of an individual developing AD. Up to 25% of Caucasian individuals are heterozygous (carry one copy) of the 4 allele, which increases the risk of developing AD by 3-fold."" 2% of the population are homozygous (carry two copies) of the €4 allele, and have a 15-fold risk of developing AD compared with the general population. More research is recommended by the Alzheimer's Association to better understand the correlation between ApoE €4 carriers and the onset of AD.° Disease-modifying therapy: Treatments or interventions that affect the underlying pathophysiology of a disease and have a beneficial outcome on the course of AD."® Biomarkers ©Institute for Clinical and Economic Review, 2023 Page Al Alzheimer's Disease - Final Evidence Report Return to Table of Contents Amyloid plaque: Extracellular deposits of the amyloid beta protein mainly in the grey matter of the brain. Beta-amyloid: Beta-amyloid (AB) plays a key role in the pathogenesis of Alzheimer's disease and can be imaged in vivo using florbetaben, florbetapir, or flutemetamol PET tracers. Beta-amyloid is formed from the breakdown of an amyloid precursor protein. Beta-amyloid refers to peptides of 36-43 amino acids that are the main component of amyloid plagques. AR monomers, a disordered peptide, can aggregate into various forms, including oligomers (soluble), protofibrils and amyloid fibrils (insoluble)." Different Alzheimer's disease therapies aim to target different species of amyloid. Amyloid plagques contain both AB40 and AB42. AB1-42 and AB1-40 are biomarkers in Alzheimer's disease detected in plasma or cerebrospinal fluid, with higher values representing more amyloid. AB1-42 is considered to be important for progression as it is more toxic. AB42/40 is also a biomarker for early detection of amyloid pathology. A lower ratio is indicative of elevated amyloid. Tau protein: A microtubule-associated protein that form insoluble filaments that accumulate as neurofibrillary tangles in Alzheimer's disease.®' Tau is suggested to help stabilize neurons in the brain and a buildup of tau impacts the function of brain cells. Tau increases with age and has been suggested to be a general marker of neurodegeneration.8! Tau is important in Alzheimer's disease as models suggest that while amyloid PET may be the earliest detected abnormal biomarker, it is followed closely by CSF tau.?? Total-tau (t-tau) is used as a key biomarker in Alzheimer's disease research. Flortaucipir: The most widely studied tau-specific radiotracer and is specific to the tau aggregates of AD. Phosphorylated tau (P-tau): Phosphorylated tau are tau in an abnormally hyperphosphorylated state. A buildup of p-tau leads to synaptic impairment and neurofibrillary tangles.®® P-tau has been considered a more specific marker for Alzheimer's disease as neurofibrillary tangles consist of tau protein in this state. Neurogranin: Calmodulin-binding protein expressed primarily in the brain, particularly in dendritic spines, and participating in the protein kinase C signaling pathway. Neurogranin is considered a biomarker for Alzheimer's disease. Plasma: The liquid component of the blood that is necessary to distribute nutrients, remove waste, and prevent infection. Plasma samples are taken in Alzheimer's disease research to examine biomarkers such as tau, p-tau, and beta-amyloid. Cerebrospinal fluid (CSF): Body fluid found in the tissue that surrounds the brain and spinal cord, providing a cushion from injury and providing nutrients to the brain. CSF samples are taken in Alzheimer's disease research to examine biomarkers such as tau, p-tau, and beta-amyloid. ©Institute for Clinical and Economic Review, 2023 Page A2 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Positron Emission Tomography (PET) scan: A minimally-invasive functional imaging technigue that uses radioactive substances (radiotracers) to visualize and measure changes in activity in organs and tissues. PET scans are used for diagnosing Alzheimer's disease by measuring the build-up of beta- amyloid in the brain. Amyloid PET is a specific type of PET scan where a ligand that binds to amyloid plaque is labeled with a radioisotope (e.g., 11C, 13N, 150, 18F) to visualize and measure amyloid. Standardized uptake value ratio {SUVr): A widely used PET quantifier. This is a method of determining activity in PET imaging and is used as a measure for quantifying the global A and tau burden. Centiloids: Standardized measure of amyloid PET imaging that scales the outcome of each analysis method or tracer to a 0 to 100 scale. The units of this scale have been named "Centiloids." SUVr can be transformed into centiloids by using individual-level data from cognitive controls and "typical" individuals with AD as anchors. Magnetic resonance imaging (MRI): Medical imaging technique used to create images of organs and tissues in the body. MRIs are used in Alzheimer's disease research to detect brain abnormalities for diagnosis and in clinical trials to monitor ARIA. Hippocampal volume: Hippocampus is a brain structure embedded in the temporal lobe that has a major role in learning and memory. Large hippocampal volume is positively associated with memory performance. Whole brain volume: Collective volume of the entire brain in structural imaging, without considering regionally specific differences in the volume of any individual structures. In Alzheimer's disease, as connections between networks of neurons break down then brain regions begin to shrink. Ventricular volume: Ventricles are one of a system of four communicating cavities within the brain and are filled with cerebrospinal fluid. Enlarged ventricles have been associated with decreases in cognitive functioning. ©Institute for Clinical and Economic Review, 2023 Page A3 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Key Outcome Instruments Cognitive Outcomes Clinical Dementia Rating - Sum of Boxes (CDR-SB) is a structured interview that assesses three domains of cognition (memory, orientation, judgment/problem-solving) and three domains of function (community affairs, home/hobbies, personal care) based on an interview with the patient or caregiver. The six domains are assigned a severity score ranging from 0 (no performance disability) to 3 (severe performance disability) and summed for a total possible score that ranges from 0 to 18.22 Higher scores suggest greater disease severity with scores between 0.5-4.0 indicating questionable cognitive impairment, 4.5-9 mild dementia, 9.5-15.5 indicating moderate dementia, and 16.0-18.0 indicating severe dementia. Minimal Clinically Important Difference (MCID) has been defined as a change of 0.98-1.63 for MCI due to AD and mild AD dementia.?" Alzheimer's Disease Assessment Scale - Cognitive Subscale (ADAS-Cog) is an objective neurological assessment that examines the severity of cognitive and non-cognitive function from mild to severe AD. Only the cognitive subscale (ADAS-Cog) was used which includes 11 subject- completed tests and observer-based assessments of memory, language, and praxis. ADAS-Cogl4 includes all 11 items plus a test of delayed word recall, number cancellation task, and a maze task with a score ranging from 0-90 with higher scores reflecting greater impairment. MCID has been defined as a change of 2 points for MCI due to AD32 and >3 points for mild AD.3940 Alzheimer's Disease Cooperative Study - Activities of Daily Living Inventory (MCI version) (ADCS- MCI-ADL) is a caregiver completed scale that measures basic and instrumental abilities. The MClI version has been adapted for individuals with mild cognitive impairment® and includes 18 items with total scores ranging from 0-53, with lower scores representing poorer functioning. There is no data on MCID. AD Composite Score (ADCOMS) is a weighted score that includes the MMSE (2 items: Constructional praxis and orientation time), CDR-SB (6 items: personal care, home and hobbies, community affairs, judgment and problem solving, orientation, and memory), and ADAS-Cogl4 (4 items: word finding difficulty, word recognition, orientation, and delayed word recall).?> Total scores range from 0-1.97. A score of <0.29 is indicative of normal cognition, 0.29 to <0.50 is indicative of MCI, 0.50 to 0.80 is indicative of mild AD, and >0.80 is indicative of at least moderate AD. There is no data on MCID. ©Institute for Clinical and Economic Review, 2023 Page A4 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Quality of Life Outcomes European Quality of Life - 5 dimensions (5 level version) (EQ-5D-5L): A self-reported visual analog scale that covers dimensions of health such as mobility, self-care, activities, pain, and anxiety/depression.8¢ Quality of Life in Alzheimer's Disease (QOL-AD): A patient-reported (or caregiver reported, e.g., subject by proxy) questionnaire reporting on quality of life in those with Alzheimer's disease.?' Zarit Burden Interview: An instrument used to measure stress experienced by caregivers of those with Alzheimer's disease.®® Other Relevant Definitions Health Improvement Distribution Index: The Health Improvement Distribution Index identifies a subpopulation that has a higher prevalence of the disease of interest and therefore, creates an opportunity for proportionately more health gains within the subpopulation. This opportunity may be realized by achieving equal access both within and outside the identified subpopulation to an intervention that is known to improve health. The Health Improvement Distribution Index is defined as the disease prevalence in the subpopulation divided by the disease prevalence in the overall population. For example, if the disease prevalence was 10% in poor Americans whereas the disease prevalence across all Americans was 4%, then the Health Improvement Distribution Index would be 10%/4% = 2.5. For interventions known to increase health in this disease and that accomplish equal access across the entire population, poor Americans would receive 2.5 times the health improvements as compared to the same sized group of Americans without regard to economic status. Health Improvement Distribution Indexes above 1 suggest that more health may be gained on the relative scale in the subpopulation of interest when compared to the population as a whole. This statistic may be helpful in characterizing a treatment's contextual considerations and potential other benefits (Section 5). For the calculation for the HIDI for African Americans, we used population estimates of clinical AD calculated by Rajan et al., which were based on the Chicago Health and Aging Project and adjusted based on the 2020 US census.?® The overall 2020 US census adjusted prevalence of clinical AD was 11.3%. e African American = 18.6%/11.3% = 1.6 ©Institute for Clinical and Economic Review, 2023 Page A5 Alzheimer's Disease - Final Evidence Report Return to Table of Contents A2. Potential Cost-Saving Measures in Alzheimer's Disease ICER includes in its reports information on wasteful or lower-value services in the same clinical area that could be reduced or eliminated to create headroom in health care budgets for higher-value innovative services (for more information, see https://icer.org/our-approach/methods- process/value-assessment-framework/). These services are ones that would not be directly affected by beta-amyloid therapies for AD (e.g., skilled nursing care), as these services will be captured in the economic model. Rather, we are seeking services used in the current management of mild cognitive impairment or mild AD beyond the potential offsets that arise from a new intervention. During stakeholder engagement and public comment periods, ICER encouraged all stakeholders to suggest services (including treatments and mechanisms of care) currently used for people with Alzheimer's disease that could be reduced, eliminated, or made more efficient. We received a suggestion that repeating neuropsychological testing to monitor cognitive decline may not be necessary. ©Institute for Clinical and Economic Review, 2023 Page A6 Alzheimer's Disease - Final Evidence Report Return to Table of Contents B. Patient Perspectives: Supplemental Information B1l. Methods ICER engaged with people with AD and caregivers, patient groups, including representatives from AD advocacy organizations and caregiver organizations, and clinical experts to gather information to better understand patient and caregiver experiences with the disease. In total, we spoke with nine persons with AD, six caregivers, two advocacy organizations, nine clinical experts, one payer, and four manufacturers via conference calls throughout the review process. We also reviewed research literature suggested by or provided to ICER by advocacy organizations as well as data from qualitative interviews and surveys of people with AD and caregivers provided to us by patient organizations.**° People living with AD, caregivers, and advocacy groups provided information on the impact of AD on individuals with the disease and caregivers throughout the disease course, particularly concerning aspects of the disease and caregiving that are not well-reflected in the current literature, and their thoughts about current and future treatment options. These informal interviews provided important information to help raise issues important to persons with Alzheimer's and their caregivers; it is important to note that the interviews were not meant to represent a formal study of patient and caregiver perspectives. Patient advocacy organizations also assisted with literature review to find information that was considered for inputs into the economic model. We spoke with a wide range of people living with AD, mainly people over the age of 50 with mild or moderate AD and mostly living at home. We spoke with both men and women, as well as both White and African American people living with AD. The caregivers we spoke with represented a wide variety of ages (from 30 to above 70) and caregiving experience, caring for loved ones across the disease spectrum (mild to severe) and in various settings (home and long-term care). We spoke with both male and female, and White, Black, and Hispanic caregivers. Clinical experts were chosen based on their expertise in diagnosing, treating, and/or researching AD, as well as recommendations from other stakeholders. We spoke with neurologists and geriatricians with expertise treating people with AD, as well as epidemiologists and clinical researchers. Clinical experts also believe that the main goal of treatment for AD is to maintain independence, and that disease-modifying drugs would be a welcome addition to the treatment arsenal. However, because there have been multiple purported disease-modifying drugs that have previously failed during the clinical trial phase, they are cautious and feel they need clear evidence ©Institute for Clinical and Economic Review, 2023 Page B1 Alzheimer's Disease - Final Evidence Report Return to Table of Contents demonstrating such an effect from a new therapy. Additionally, clinical experts also cited issues with clinical trial outcome measures, including that CDR-SB may not be sufficiently sensitive to changes in cognition and function. Payers were concerned about the use of surrogate markers as endpoints in clinical trials, and the lack of clear association between those surrogate markers and clinical outcomes. Manufacturers noted the urgency of developing new treatments for AD, the disproportionate impact of AD on persons of color and their caregivers, and that better information on the impact of AD treatments on quality of life for both individuals with the disease and caregivers is needed. ©Institute for Clinical and Economic Review, 2023 Page B2 Alzheimer's Disease - Final Evidence Report Return to Table of Contents C. Clinical Guidelines Clinical practice guidelines for the treatment of MCl and mild AD have been issued by several US and non-US-based organizations. These guidelines are summarized below. American Academy of Neurology® In 2018, the American Academy of Neurology published guidelines for the management of MCI. The guidelines recommended that clinicians assess for MCl using validated tools, evaluate individuals with MCI for modifiable risk factors, assess for functional impairment, assess for and treat behavioral symptoms, and consider discontinuing medications that may impair cognition. Furthermore, guidelines suggested that clinicians should counsel patients about the expected course of the disease, encourage long-term planning, and discuss the lack of effective medication options, including the lack of benefit of cholinesterase inhibitors on cognition and progression. National Institute for Health and Care Excellence (NICE)® Guidelines for the diagnosis and management of dementia were published in June 2018 by NICE in the United Kingdom. The guidelines include recommendations on involving people living with dementia in decisions about their care, assessment and diagnosis of dementia, interventions to promote cognition, independence and well-being, pharmaceutical interventions, managing non- cognitive symptoms, supporting caregivers, and staff training and education. Among the non- pharmacological interventions recommended were group cognitive stimulation and reminiscence therapy and cognitive rehabilitation, and recommendations against acupuncture, herbal supplements, vitamin E, and non-invasive brain stimulation. Consideration should be given to minimizing medications that may impair cognition. Acetylcholinesterase inhibitors were recommended for managing mild-to-moderate AD symptoms, and memantine and/or combination therapy was recommended for moderate-to-severe AD. Recommendations were also made to manage non-cognitive symptoms (e.g., behavioral symptoms, depression, sleep problems), and managing other long-term conditions common in people with AD, such as pain, falls, and incontinence. American Psychiatric Association" The American Psychiatric Association published practice guidelines for the treatment of individuals with AD in 2014. The guidelines discuss the evidence of efficacy for medications to treat AD, and state that based on the available evidence, memantine, cholinesterase inhibitors, or a combination of the drugs, may be used to treat AD. They also recommend using nonpharmacological interventions and environmental measures to reduce psychosis and agitation before considering use of antipsychotics based on the lack of evidence for efficacy of antipsychotics in this situation. ©Institute for Clinical and Economic Review, 2023 Page C1 Alzheimer's Disease - Final Evidence Report Return to Table of Contents The guidelines also discuss the evidence for a variety of psychosocial interventions and alternative treatments, and offer guidance on managing caregiver stress. The National Institute on Aging-Alzheimer's Association'374%2 In 2011, the National Institute on Aging and the Alzheimer's Association convened a workgroup to revise the diagnostic criteria for MCl and AD. These included diagnostic criteria both to be used in the clinical setting and in research settings. Clinical and cognitive criteria were established to differentiate MCl and AD, and to establish the potential etiology of MCIl. Furthermore, for AD, diagnostic criteria incorporating biomarkers were defined. Biomarkers to incorporate into research criteria were also discussed, including PET amyloid imaging for beta-amyloid deposition and CSF fluid tau/phosphorylated tau, among others. In 2018, the National Institute on Aging and the Alzheimer's Association issued an updated research framework intended to guide observational and interventional research®2. The objective was to create a scheme for defining and staging AD across the lifespan. The framework establishes a biomarker-based system for classifying the neuropathologic changes seen in AD, including imaging and CSF biomarkers. Biomarkers are separated into those related to beta-amyloid plaques (e.g., CSF AB-42, amyloid PET), fibrillar tau (e.g., CSF phosphorylated tau, tau PET), and neurodegeneration or neuronal injury {e.g., anatomic MRI, total CSF tau). Categorization of AD- and non-AD-related pathologic change using biomarkers is discussed. Additionally, the document discusses cognitive staging applicable to research cohorts, including syndromal categorical cognitive staging that uses traditional syndromal categories {cognitively unimpaired, MCl, dementia), and numeric clinical staging (from Stage 1 cognitively normal to Stage 6 severe dementia) for people with the disease in the AD continuum. Aducanumab: Appropriate Use Recommendations® After FDA approval of aducanumab, a panel of Alzheimer's disease experts convened to discuss best practices for aducanumab use. The panel defined the appropriate patient population for initiation of treatment, which includes individuals with early AD with positive amyloid via PET scan or CSF investigation, and recommends potential exclusion criteria, including concomitant use of anticoagulation. The expert panel also recommend an MRI monitoring schedule of baseline and surveillance MRIs based on aducanumab titration, and management strategies for ARIA, including MRI monitoring intervals and recommendations for dosing. Finally, the expert panel made recommendations on the healthcare system resources needed for safe and appropriate use of aducanumab, such as clinical expertise, access to testing for amyloid, infusion resources, access to MRI, and access to family and patient education and support. ©Institute for Clinical and Economic Review, 2023 Page C2 Alzheimer's Disease - Final Evidence Report Return to Table of Contents D. Comparative Clinical Effectiveness: Supplemental Information D1. Detailed Methods A prior version of this report included a review of donanemab. Due to the manufacturer receiving a Complete Response Letter from the US Food and Drug Administration (FDA) on January 19, 2023 for donanemab's accelerated approval biologics license application, we removed it from the review, starting with the Revised Evidence Report. There was no new clinical evidence for aducanumab since the previous August 2021 review?® that would have affected our assessment and thus aducanumab was not reviewed in this report. PICOTS Population The population of focus for the review was adults with early AD (i.e., MCI due to AD [also termed I" "prodromal" Alzheimer's] and mild AD dementia) with evidence of AD pathology (e.g., amyloid positivity). This population approximates individuals whose condition would be categorized as Stages 3 or 4 using diagnostic criteria outlined by the FDA.** Evidence that includes individuals with AD in Stage 2 will only be considered if the sample also includes individuals in Stage 3. Interventions The intervention of interest for this review was lecanemab in addition to supportive care. Supportive care includes both non-pharmacologic and non-disease-modifying pharmacologic interventions. Comparators We compared lecanemab in addition to supportive care to supportive care alone. ©Institute for Clinical and Economic Review, 2023 Page D1 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Outcomes The outcomes of interest are described in the list below. ©lnstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Patient-centered Outcomes o Changein: O O O 0O 0O O O O = Ability to maintain independence and autonomy = Ability to perform activities of daily living (e.g., as measured by AD Cooperative Study-Activities of Daily Living Inventory-MClI, etc.) = Cognitive function (e.g., as measured by Clinical Dementia Rating, Mini- Mental State Examination, AD Composite Score, Alzheimer's Disease Assessment Scale - Cognitive Subscale, Integrated Alzheimer's Disease Rating Scale, Montreal Cognitive Assessment Test, etc.) = Neuropsychiatric symptoms (e.g., as measured by Neuropsychiatric Inventory Questionnaire) Delayed entry into institutional care Disease progression Symptom progression Maintenance of identity and personality Quality of life Emotional wellbeing Caregiver impact = Caregiver quality of life = Caregiver health = Caregiver productivity Behavioral change Ability to communicate Adverse events including but not limited to = Serious adverse events = Discontinuation due to adverse events = Infusion-related reactions = Death = Symptomatic amyloid-related imaging abnormalities (ARIA-E and ARIA-H) Other Outcomes Level of beta-amyloid (e.g., PET, CSF) = Percentage of amyloid e Percentage reduction e Absolute percentage = Amyloid clearance e Mean reduction in amyloid from baseline Page D2 Return to Table of Contents e Percentage of participants reaching amyloid negativity e Rapidity of participants reaching amyloid negativity = Durability of biomarker reductions (e.g., tau levels and beta-amyloid) Level of tau proteins (e.g., CSF phosphorylated tau, total tau, PET ligand) Neuroinflammation Brain atrophy Brain volume (e.g., hippocampal volume, ventricular volume, or whole brain volume) o Additional biomarkers may be reviewed based on input from manufacturers and O O O O clinical experts as the review progresses Timing Evidence on intervention effectiveness and evidence on harms were derived from studies of any duration. Settings Evidence from all relevant settings was considered with a particular focus on the outpatient setting. ©Institute for Clinical and Economic Review, 2023 Page D3 Alzheimer's Disease - Final Evidence Report Return to Table of Contents S1U9U0) JO 2|qe] O} uiniay Q 98ed 1loday a3uapiag |eul{ - aseasiq S, JawWidyz|y €707 'M3IASY J1WOUODT pue [edjul|] 10} SINISUIQ 'sisAjeue-e1aw yoea Joj (| QIEHEEVRET) ¥T S}NS3J JO SISOYIUAS ~8-3) Aduaisisuod Jo saunseaw Suipnppul 'auop Jl "sa1pnis Jo s3NsaJ ujuiquiod pue elep Sujjpuey Jo Spoylaw 3yl aquasag '(sueaw u) S0uaJBYIp 'Oes s "3'3) saunseaw Asewwns |edpuiid syl Ieis €T sainseaw Ajewwinsg *SIS9Y3uAs ejep Aue uj pash aq 03 S| UOJIBWIOUI SIY} MOY pue '([]9A3] wod3no Jo Apnis 3y} saIpnis 1€ dUOp Sem SIY} JaY1aym Jo uonesydads Suipnoul) salpnis [BNPIAIPUL JO SBIq JO YS1 BulSSasse o) pash spoylaw aquasag ct |ENPIAIPUI Ul SBIq JO )SIY "'9pew suojeayidwis 1T sway ejeq pue suondwnsse Aue pue (s924nos Suipuny 'SOJId "8°3) IY3nos a1om e1ep Ydlym Joj SS|geLIeA [|B SUljap pue 1si] 's103e813s9Aul wody eyep Sujwaijuod pue Sujuieiqo OT | ssad0ud uonaa|od eleq Joy} sassado.d Aue pue (ajedldnp uj 'Ajpuspuadapul 'swuoy pajojid "8'a) suodas wouy uondeIxo elep JO poyidw aqlasad *(s1sAjeue-e1ow ay3 ul 6 uoI1193|3s Apnis papnjaul 's|gedidde Ji 'pue 'mainaJ d13ew3sAs ul papnpoul 'Ajiqi8ie 'Buiuaauds 9°1) saipnis 8ul3d9|9s 404 sso20.4d By} 31e1S 'paeadal aq pinod H 18yl Yyons 'pasn sywi| Aue Suipnpul 'Oseqejep auo jsed| je 10} ASd1eJ)s Yoaeas J1u041I3|d [N} JUISAId 8 youeas "pPay2Jeas 1se| ep pue yd4eas ay3 ul (saipnis Py 18] 93Ep pUe Y ua ut (satpn3 L $92JN0S uollew.ou| |euonnippe Aj1auapi 03 sioyine Apnis yum 10e3uod '98esan0d Jo sai1ep Ylim saseqelep "8:9) s904n0S UoleWIOLUI [|B 3qLI2saQ *9jeuoiies Suiai8 'A1|1qi8iS Joy elI911UD Se pasn (sn1els uonediqnd 9 els10 Anjiqi83 'aBen3due| 'paJopisuod siedA "3'9) sanisiualoeseyd Hodas pue (dn-mojjo4 Jo Yyi8ua) 'SODId "8'9) solisiardeIeyd Apnis Ajlaads *Jaqwinu uoileJ1si8aJ Suipn[oul UolleWIOUl UOIIRIISISD uoijessi8al apinoud 'S|gejieAe Ji 'pue '(ssauppe g "8'9) passa03e 90 UBD 3 SJ3YM pue I 'SISIX |020304d MIIASJ € §I 93edipu| : pue |02030.d SAOH13IN *(SODId) udisap Apnas pue 'saw023no v SaAI03(q0 'suosiiedwod 'suojluanialul 'syuedidizied 03 32U49434 YUM passalppe Sulaq suolrsanb Jo Juswaiels 1didxs ue apinoid 'umouy| ApesJje s| Jeym JO 1X31U03 3Y3 Ul MBIASJ BY] J0j B|euojiel ay3 2qL1asaq € 9jeuoney NOILONAOYLNI -19quinu uoj1el1s|8a1 malndl o1ewdlsAs [s8uipuly Ay Jo suoieaidwi pue suoisnjpuod 'suoijenw| {s3nsaJ 'spoylaw sisayjuAs pue |esiesdde Apnis {suonusAialul pue 'syueddiyed Z Alewwins painionuls 'ela)ud ANjIgiSie Apnis {s224nos elep {saAlldalqo 'punoidydeq :a|qedijdde se 'Guipnjoul Alewiwins paJnianuls e apiAo.d 1ovy1sav *yzoq Jo 'sisAjeue-elaw 'Mmalaal d11ewalsAs e se Jodad syl Ajlauap) _ T _ S[IL LI ISIP1I3YD 600Z VINSINC "TA 3Iqel SJUaIUO) JO 9|ge 0} uIN1ay 1oday 27UapIAT [euUl{ - seasI( S JaWIsYyzZ|y 5Q 98ed €707 'M3IASY J1WOUODT pue [edjul|] 10} SINIISUIQ £60000TPawWd"[euInof/TLETOT:I0P "L60000T3 :(9)9 P3IAl SO1d "1UBWSIEIS VINSIYd 9y :sasAjeuy-e19|A] pue SMIADY d11eWISAS 10y swa)| Sunuoday palisiald (600Z) dnoio YIASIHd 9YL "D Uewl|y 'T H4e|z12] 'v 11e1aqi] 'g 19Yyow :wold *M3IA3J J13eWR)SAS 8uipun 9yz1 Joy s1apuny Jo ajou (elep jo Ajddns ""8'a) 1ioddns Jaylo pue malaal d11ewWI1SAS BYy] 40} Suipuny Jo $921N0S 3qLIISAQ] Le fpund ONIANNS "Yo1edsaJ 913Ny 10} suojled]jdw] pue '93UdpIA JSYIO0 JO IXSIU0D dY) U SHNSI 3y} Jo uoielaidialul [esauas e apinold | 9g SuUoISNPUO) *(seiq Suiuodau 'youeasald S¢ suoljejwn pal}11uapl Jo [eAd1I3ad 339|dwodUl "5°3) [9A3]-M3IAS 1E pue '(selq Jo jsi "8'3) [9A3] W 033N0 pue Apnis 3e suoljeliwi| ssnasig *(s19yew Adjjod pue 'siasn 'siapiaoid aseayyesy "8-a) sdnoas ¥z 92U3PIAS JO Alewwing A3y 01 3dUBAB[2J JI9Y} J3PISUOD {3LU0JIN0 UlRW Yoea Jo) 32UPIA3 Jo Yidualls ayl Suipnjpul s8uipuly ulew ay) aziewwns NOISSNoSId ([9T way| 99s] uoissaiBal-e1aw 'sasAjeue dnoidqns Jo AlAIISUDS "5'9) Suop JI 'sasAjeue [euollippe JO SNSAJ AAID | €7 sishjeue jeuonippy salpnis '(ST wa3| 93S) SaIPN3S SSOJIE Skl JO Sl JO JUBWISSDSSe Aue Jo S}nsaJ Juasald | ¢ L SsoJde selq Jo dsiy *A2u3315I5U0J JO SAUNSEBW PUEB S|BAIDIU] S2UIPLUOD Bulpn|du] 'BUop SiSA|eue-BIaW YIBa JO S}NSSJ JUdsald | T2 S} NsaJ JO SISBYIUAS -30|d 153404 B yM Aj|eSpI 'S|EAIS1UI DIUDPIIUOD pUB S91BWILSS 193443 (q) dnoud SaIpn3s UOoI3UBAJIRIUI Yyoea Joy ezep Asewwns ajdwis (e) :Apnis yoea Joj 'quasald '(swuey Jo SsHyauaq) pa4apISUOI SBWOIINO ||e J04 0c |eNpPIAIPUL JO S} NS3Y salpnis ° WD)l 99S) JUSWISSISSE [9AD| 2W021N0 Aue 'a|ge|ieAe J1 'pue Apnis Yoea Jo Seiq JO )SI1 Uo elep Juasal (2T way 3935)1 [oA3| 2W00] qejiene 41 'pue Apn3s yoea 4O Selq 4O s Uo eyep Juasald | 6T LI eI 40 Y1 'suoned syl 8T sa11sl1a12eIeYd ApNnls apinoud pue (porad dn-mojjo) 'SODId '©z1s Apnis "8°9) paldelixs a4aM el1ep Ydiym Joy sallsiialoeleyd jussald 'Apnis yoes Jo4 'weJsSe|p Mo} e yum Ajjeapi 'a08els LT uo3as Apnis yoea 1. SUOISN|oXa JOJ SUOSEIJ YUM 'M3IAaJ 3Y] Ul papnaul pue 'A11|1qi81|9 Jo) passasse 'Pauaasds salpnis JO SIaquinu aAID S11NS3y 'payads-aud asam 9T sasA|eue |euonippy yoiym Buneoipul 'suop y '(uoissaidau-elaw 'sasAjeue dnoudqns 1o AlAILSUSS "8°9) sasAjeue |euonlippe JO Spoylaw aquIdsa( '(s21pN31s ulyHM salpnis 8uiodal aAnd3|as 'seiq uonediqnd "3'9) 92uUapIAS SAIIR|INWND BY3 1094e Aew 1eyl selq Jo YISk Jo Juawissasse Aue Ajoads st SsoJoe selq Jo ysiy QI EETVRE ) Data Sources and Searches Procedures for the systematic literature review assessing the evidence on donanemab, lecanemab, and updated evidence for aducanumab, in addition to supportive care, for the treatment of early Alzheimer's disease (AD), i.e., MCl due to AD and mild AD dementia followed established best research methods.>>% We conducted the review in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.'" The PRISMA guidelines include a checklist of 27 items, which are described further in Table D1. We searched MEDLINE, EMBASE, Cochrane Database of Systematic Reviews, and Cochrane Central Register of Controlled Trials for relevant studies. Each search was limited to English-language studies of human subjects and excluded articles indexed as guidelines, letters, editorials, narrative reviews, case reports, or news items. All search strategies were generated utilizing the Population, Intervention, Comparator, and Study Design elements described above. The proposed search strategies included a combination of indexing terms {MeSH terms in MEDLINE and EMTREE terms in EMBASE), as well as free-text terms. At the time of the updated search (02/02/2023), lecanemab had been given a brand name {Leqembi) and, for completeness, this search term was included in the updated search. To supplement the database searches, we performed a manual check of the reference lists of included trials and reviews and invite key stakeholders to share references germane to the scope of this project. We also supplemented our review of published studies with data from conference proceedings, regulatory documents, and information submitted by manufacturers, and other grey literature when the evidence met ICER standards (for more information, see https://icer.org/policy- on-inclusion-of-grey-literature-in-evidence-reviews/). Where feasible and deemed necessary, we also accepted data submitted by manufacturers "in-confidence," in accordance with ICER's published guidelines on acceptance and use of such data (https://icer.org/guidelines-on-icers- acceptance-and-use-of-in-confidence-data-from-manufacturers-of-pharmaceuticals-devices-and- other-health-interventions/). ©Ilnstitute for Clinical and Economic Review, 2023 Page D6 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Table D1.1. Search Strategy of Medline 1996 to Present with Daily Update and Cochrane Central Register of Controlled Trials (lecanemab and donanemab) {lecanemab or 'ban 2401' or ban2401 or Legembi).ti,ab. {donanemab or 'ly 3002813' or ly3002813).ti,ab. lor2 AIWIN(- {addresses or autobiography or bibliography or biography or comment or congresses or consensus development conference or duplicate publication or editorial or guideline or in vitro or interview or lecture or legal cases or legislation or letter or news or newspaper article or patient education handout or periodical index or personal narratives or portraits or practice guideline or video audio media).pt. 3not4 {exp animals/ or exp animal/ or exp nonhuman/ or exp animal experiment/ or animal model/ or animal tissue/ or non human/ or (rat or rats or mice or mouse or swine or porcine or murine or sheep or lambs or pigs or piglets or rabbit or rabbits or cat or cats or dog or dogs or cattle or bovine or monkey or monkeys or trout or marmoset$1 or basic research or cell lines or in vitro or animal model or canine).tw.) not {humans/ or human/ or human experiment/ or (human* or men or women or patients or subjects).tw.) 5not6 Limit 7 to English Language Remove duplicates from 8 Table D1.2. Search Strategy of EMBASE SEARCH (lecanemab and donanemab) #1 | lecanemab/exp #2 | (lecanemab OR "ban 2401" OR ban2401 OR Leqembi):ti,ab #3 | #1 OR#2 #4 | donanemab/exp #5 | (donanemab OR "ly 3002813" OR ly3002813):ti,ab #6 | #4 OR#5 #7 | #3 OR #6 #8 | ('case report'/de OR 'practice guideline'/de OR 'questionnaire'/de OR 'chapter'/it OR 'conference review'/it OR 'editorial'/it OR 'letter'/it OR 'note'/it OR 'short survey'/it) #9 | #7 NOT #8 #10 | ('animal'/exp OR 'nonhuman'/exp OR 'animal experiment'/exp) NOT 'human'/exp #11 | #9 NOT #10 #12 | #11 AND [english]/lim ©Institute for Clinical and Economic Review, 2023 Page D7 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Table D1.3. Search Strategy of Medline 1996 to Present with Daily Update and Cochrane Central Register of Controlled Trials (updated evidence for aducanumab) 1 | (aducanumab or BIIBO37 or 'BIIB 037' or BIIB-037 or BIIB37 or BlIB-37 or aduhelm).ti,ab. 2 | (addresses or autobiography or bibliography or biography or comment or congresses or consensus development conference or duplicate publication or editorial or guideline or in vitro or interview or lecture or legal cases or legislation or letter or news or newspaper article or patient education handout or periodical index or personal narratives or portraits or practice guideline or review or video audio media).pt. 3 1NOT2 4 | (exp animals/ or exp animal/ or exp nonhuman/ or exp animal experiment/ or animal model/ or animal tissue/ or non human/ or (rat or rats or mice or mouse or swine or porcine or murine or sheep or lambs or pigs or piglets or rabbit or rabbits or cat or cats or dog or dogs or cattle or bovine or monkey or monkeys or trout or marmoset$1 or basic research or cell lines or in vitro or animal model or canine).tw.) not (humans/ or human/ or human experiment/ or (human* or men or women or patients or subjects).tw.) 3NOT4 Limit 5 to English Language Remove duplicates from 6 Limit 7 to yr="2020 -Current" || (U Table D1.4. Search Strategy of EMBASE SEARCH (updated evidence for aducanumab) #1 | 'aducanumab/' #2 | aducanumab:ti,ab OR biib037:ti,ab OR 'biib 037':ti,ab OR biib37:ti,ab OR 'biib-37":ti,ab OR aduhelm:ti,ab #3 | #1 OR#2 #4 | #3 NOT ('case report'/de OR 'human tissue'/de OR 'nonhuman'/de OR 'practice guideline'/de OR 'questionnaire'/de OR 'chapter'/it OR 'conference review'/it OR 'editorial'/it OR 'letter'/it OR 'note'/it OR 'review'/it OR 'short survey'/it) #5 | {'animal'/exp OR 'nonhuman'/exp OR 'animal experiment'/exp) NOT 'human'/exp #6 | #4 NOT #5 #7 | #6 AND [english]/lim #8 | #7 AND [18/05/2021]/sd ©Institute for Clinical and Economic Review, 2023 Page D8 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Figure D1. PRISMA flow Chart Showing Results of Literature Search for Aducanumab, Lecanemab and Donanemab for Alzheimer's Disease. 197 references identified 22 references identified through literature search through other sources 45 references after duplicate removal 134 citations excluded 23 citations excluded 40 references assessed for 4 Population eligibility in full text 16-Study Design 2 Outcome Additional Data 14 conference abstracts and 9 Donanemab citations presentations removed* 7 Data submissions *At the time of our revised report, donanemab had been removed as an intervention from this report. Thus, nine references that provided data on donanemab were removed. TOur search did not identify any new aducanumab references. Thus, aducanumab was not included in our review. Study Selection We performed screening at both the abstract and full-text level. A single investigator screened all abstracts identified through electronic searches according to the inclusion and exclusion criteria described earlier. We did not exclude any study at abstract-level screening due to insufficient information. For example, an abstract that did not report an outcome of interest would be accepted for further review in full text. We retrieved the citations that were accepted during abstract-level screening for full text appraisal. One investigator reviewed full papers and provided justification for exclusion of each excluded study. ©Institute for Clinical and Economic Review, 2023 Page D9 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Data Extraction Data were extracted into Microsoft Excel. The basic design and elements of the extraction forms followed those used for other ICER reports. Elements included a description of patient populations, sample size, duration of follow-up, funding source, study design features, interventions (agent, dosage, frequency, schedules), concomitant therapy allowed and used (agent, dosage, frequency, schedules), outcome assessments, results, and risk of bias for each study. The data extraction was performed in the following steps: 1. One reviewer extracted information from the full articles, and a second reviewer validated the extracted data. 2. Extracted data were reviewed for logic, and a random proportion of data were validated by a third investigator for additional quality assurance. Risk of Bias Assessment We examined the risk of bias for each randomized control trial in this review using criteria published in the Cochrane Risk of Bias Assessment Tool Version 2.%%%9 Risk of bias was assessed by study outcome for each of the following aspects of the trials: randomization process, deviation from the intended interventions, missing outcome data, measurement of the outcome, selection of the reported results, and overall risk of bias. Two reviewers independently assessed these domains. Any disagreements were resolved through discussion or by consulting a third reviewer. We did not assess the risk of bias in trials where we only had access to conference abstracts/presentations. " " To assess the risk of bias in trials, we rated the categories as: "low risk of bias", "some concerns", or "high risk of bias". Guidance for risk of bias ratings using these criteria is presented below: Low risk of bias: The study is judged to be at low risk of bias for all domains for this result. Some concerns: The study is judged to raise some concerns in at least one domain for this result, but not to be at high risk of bias for any domain. High risk of bias: The study is judged to be at high risk of bias in at least one domain for this result or the study is judged to have some concerns for multiple domains in a way that substantially lowers confidence in the result. We examined the risk of bias for the following outcomes: CDR-SB, ADAS-Cog14, ADCS-MCI-ADL, and ARIA. See Table D1.5. ©Institute for Clinical and Economic Review, 2023 Page D10 Alzheimer's Disease - Final Evidence Report Return to Table of Contents SJU9UO) JO 3|qe] 0} uIN}ay 1TQ @3ed 1oday a3uapiAg |eul{ - asessiq S, JawWidyz|y €707 'M3IASY J1WOU0DT pue [edjul|] 10} SINISUIQ saxog JO wns - Suiley eluswaq [eaul :gS-yad 'saiijewsouqe SujSew) palejal-piojAwy (yyy '(UoiISIaA [DIN) Alojuanu) Suia Ajleq jo sa1lARY - Apnis aAjleladoo) aseasiq S JawiBYzly 1av-1DIN-SDAV '31eosqns aAIHUS0) - 5]edS JUBWISSASSY aseasiq S Jawiayzly (pT80d-svav dnoJ8 |eyuswiadxa ay) paJoaey ysH ysdiy,, 40 ,SUIDIUOD SWOS,, ||V 210N YsiH Mo ysiH Mo Mo MO oe AV ALIYVTD viuy SUJIIUO0D WIS MO SUJ9IU0I WO Mo MO MO oc AV ALIIVTD 1av-PDN-Sdav SUJ3IUO0I BWOS MO MO MO SUJI3OU0D BWOS MO og AV ALIYVTD p1803-SVAY Mo MOT SUJ92U0d dWoS Mo MO MO oc AV ALIVTD gs-4ad sel nso 9110dd Wwodln0 © eje suoRuaAIu| $S930.1 0 )SI .mm._w> _._.w_ M"o"__. J9| wm_ 0 :mfiu__.mhnw_mm-_" 2002 :u M:_mm_ papuaiu) ay3 uol NN_Eon_:m swieu jellL JOSIY [|e4AN0 43 JO Uo1}IB|3S J03 W O Suissi wouy uopeINaq 1jezjiwopuey SJUBISSASSY Seld JO Sy *S°Ta d|qel Evaluation of Clinical Trial Diversity We evaluated the demographic diversity of clinical trials using the ICER-Developed Clinical Trial Diversity Rating Tool. The tool is designed to evaluate the three demographic characteristics described in Table D1.6. below. Additional guidance on using the tool is described below. Table D1.6. Demographic Characteristics and Categories Demographic Characteristics Categories 1. Race and Ethnicity Racial categories: e White e Black or African American e Asian e American Indian and Alaskan Native e Native Hawaiian and Other Pacific Islanders Ethnic Category: e Hispanic or Latino 2. Sex e Female s Male 3. Age e QOlder adults (265 years) Rating Guide Representation for each demographic category is evaluated relative to the disease prevalence, using the metric "Participant to Disease-prevalence Representation Ratio" (PDRR). Next, a representation score is assigned based on the PDRR estimate. The score for each demographic category ranges from O to 3 based on the PDRR cut points presented in Table D1.7. Finally, based on the total score of the demographic characteristics (e.g., race and ethnicity), the categories "good," "fair," or "poor" are used to communicate the overall level of diversity of a clinical trial. The description of the rating categories for each demographic characteristic is provided in Table D1.8. Asecond diversity rating that evaluates the subpopulation of patients recruited from the US is provided for multinational trials. Table D1.7. Representation Score PDRR Score 0 or Information on Demographic Category Not Reported >0 and Less Than 0.5 0.5t0 0.8 20.8 PDRR: Participant to Disease-prevalence Representation Ratio WIN|=]O ©Institute for Clinical and Economic Review, 2023 Page D12 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Table D1.8. Rating Categories Demographic . . Maximum . . . . Demographic Categories Rating Categories (Total Score) Characteristics Score Good (11-12 .. Asian, Black, White, and oo. ( ) Race and Ethnicity* . . 12 Fair (7-10) Hispanic Poor (<6) Good (6) Sex Male and Female 6 Fair (5) Poor (<4) Good (3) Age Older adults (265 years) 3 Fair (2) Poor (<1) *American Indian or Alaskan Native & Native Hawaiian or Other Pacific Islander are not factored into the overall racial and diversity rating. However, information on enrollment and PDRR estimates are reported when reliable prevalence estimates are available. Results Table D1.9. Race and Ethnicity White Black Asian Hispanic Total Diver.'sity AIAN NHPI score Rating Prevalence 72.10% | 16.97% 4.55% 20.71% 1.09% NR G000-201 NR NR NR NR - - NR NR PDRR NC NC NC NC - - NC NC Score | 0 0 0 0 0 Poor -- -- CLARITY AD 76.90% | 2.60% 16.90% 12.90% - - 0.11% 0.06% PDRR | 1.07 0.15 3.71 0.6 - - 0.10 NE Score | 3 1 3 2 9 Fair -- -- AIAN: American Indian or Alaskan Native, NR: Not Reported, NC: Not Calculated, NE: Not Estimated, NHPI: Native Hawaiian or Pacific Islander, PDRR: Participant to Disease-prevalence Representation Ratio ©Institute for Clinical and Economic Review, 2023 Page D13 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Table D1.10. Race and Ethnicity in US Participants Only White Black Asian Hispanic Total Divefsity AIAN NHPI score Rating Prevalence 72.10% 16.97% 4.55% 20.71% 1.09% NR G000-201 NR NR NR NR - - NR NR PDRR NC NC NC NC - - NC NC Score - - - - - - - - CLARITY AD 94.80% | 4.40% 0.80%* 21.90% - - NR NR PDRR | 1.31 0.26 0.18 1.06 - - NC NC Score | 3 1 1 3 8 Fair - - AIAN: American Indian or Alaskan Native, NR: Not Reported, NC: Not Calculated, NHPI: Native Hawaiian or Pacific Islander, PDRR: Participant to Disease-prevalence Representation Ratio *We estimated the number of Asian participants in the US based on the subgroup analyses of US participants reported in the van Dyck et al. (2022) manuscript. However, these participants likely included those who were AIAN or NHPI; thus, the number of Asian participants was likely lower than our estimation. Table D1.11. Sex and Age Sex Age Male Female Score Rating || Older Adults (265 years) Score Rating Prevalence 38.40% | 61.60% NR G000-201 50.4% | 49.60% - - - - - PDRR | 1.31 0.81 - - NC - - Score | 3 3 6 Good - NC NC CLARITY AD 47.70% | 52.30% - - - - - PDRR | 1.24 0.85 - - NC - - Score | 3 3 6 Good - NC NC NC: Not Calculated, PDRR: Participant to Disease-prevalence Representation Ratio Assessment of Level of Certainty in Evidence We used the ICER Evidence Rating Matrix to evaluate the level of certainty in the available evidence of a net health benefit among each of the interventions of focus (see Appendix D).100,10 Assessment of Bias As part of our quality assessment, we evaluated the evidence base for the presence of potential publication bias. Given the emerging nature of the evidence base for these newer treatments, we scanned the ClinicalTrials.gov site to identify studies completed more than two years ago. Search terms included lecanemab, ban2401, donanemab, ly3002813, aducanumab, aduhelm, and biib037. We selected studies which would have met our inclusion criteria, and for which no findings have been published. We will provide qualitative analysis of the objectives and methods of these studies Page D14 Return to Table of Contents ©lnstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report to ascertain whether there may be a biased representation of study results in the published literature. Data Synthesis and Statistical Analyses The studies were summarized in the text and in evidence tables of the Evidence Report. This summary is key to understanding the evidence base pertaining to the topic. Any key differences between the studies in terms of the study design, patient characteristics, interventions (including dosing and frequency), outcomes (including definitions and methods of assessments), and study quality was noted in text of the report. D2. Additional Clinical Evidence Lecanemab CLARITY AD Methods The inclusion criteria for CLARITY AD were participants aged between 50 and 90 years, had a mild cognitive impairment due to Alzheimer's disease or mild Alzheimer's disease dementia, objective impairment in episodic memory, amyloid positive as determined by PET or CSF (AB1-42), had Mini- Mental State Exam (MMSE) score of 22-30, BMI 17-35, and, if they were receiving any treatment for AD symptoms, were on a stable dose for 12 weeks prior to baseline. Participants were excluded if they had any other neurological condition that may be contributing to cognitive impairment, any psychiatric diagnosis/symptoms, Geriatric Depression Scale (GDS) score =8, contraindications to MRI, or lesions on MRI that could indicate dementia diagnosis other than Alzheimer's disease. Full inclusion and exclusion criteria are described in Table D3.1. in the Supplement. During the trial, an amendment was made to the information sheet to highlight the risk of brain bleeding.*' Additional Results Cognitive Outcomes Prespecified subgroup analyses were conducted on the cognitive outcomes. In the main report, we described four subgroups that were consistent across the four cognitive outcomes. In this supplement, we also present the race (global) subgroup. Black participants showed a greater slowing of decline compared to White and Asian participants. See Figure D2. A greater slowing of decline in Black participants was also seen in ADCOMS and ADCS-MCI-ADL. See Supplement Tables D3.20-21. However, this group consisted of 44 individuals and thus should be interpreted with caution. ©Institute for Clinical and Economic Review, 2023 Page D15 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Figure D2. Forest Plot of Subgroup Analyses for CDR-SB at 18 Months {Age, ApoE €4 Genotype, Sex, and Race) % Mean difference Subgroup Slowing in CDR-SB ApoE4 non-carrier 41 -0.75 A ApoE4 heterozygote 30 05 -- ApoE4 homozygote 22 0.28 s Female 12 02 S Male 43 0.73 -- <B5 Years 6 -0.08 P 65-74 Years 23 037 -a- >=T75 Years 40 0.72 -- White 27 -0.49 - Asian 19 -0.35 i Black 63 -0.72 f = i Beta-amyloid Levels Changes in beta-amyloid are described in the main report. Changes in plasma AB42/40 ratio and CSF values (AB1-42 and AB1-40) are described in this supplement. For participants who consented to the CSF substudy, CSF samples were taken at baseline, week 53, and week 77, and plasma samples were taken at baseline, week 53, and week 79. Based on an analysis with 73.3% of the total participants (N=1316), plasma AB42/40 ratio increased more in the lecanemab group (0.008), compared to the placebo group which remained constant (0.001). Based on an analysis with 11% of the total participants (N=198) participants, there was an increase in CSF AB1-42 in the lecanemab group (281.8), compared to a very slight decrease in the placebo group (-6.5). Based on an analysis with 7.9% of the total participants (N=142), there was a decrease in CSF AB1-40 in the lecanemab group (-417.9), compared to a decrease in the placebo group (-89.9).*3 The investigators noted that these findings suggest sustained amyloid reversal effects. ©Institute for Clinical and Economic Review, 2023 Page D16 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Brain volume was assessed using MRI. MRI scans were taken at screening and week 9, 13, 27, 53, and 79 (or at early termination visit). Volumetric MRI data was available for approximately 75% of participants in the CLARITY AD trial at 18 months, and showed increased atrophy in the whole brain and cortical thickness, and decreased atrophy in the hippocampus. Data was digitized from the figures presented at the Clinical Trials on Alzheimer's Disease {CTAD) 2022 conference*® and are reported in Supplement Table D3.4. Biomarkers T-tau was examined in 199 participants consenting to the tau substudy. Tau was examined by CSF at baseline, week 53, and week 77 and, in those consenting to the PET substudy, via PET at baseline, week 59, and week 79. There was a significant decline in CSF tau in the lecanemab group, compared to an increase in the placebo group (-29.1 vs 95.3, p<0.001).3 Subgroup analysis conducted examining changes in tau pathology, as measured by PET, across different brain regions showed that the mean difference in tau pathology was significant in the medial temporal (-0.068, p=0.002), meta temporal (-0.07, p=0.012), and temporal areas {-0.065, p=0.016).** See Supplement Table D3.4. P-tau was examined by CSF at baseline, week 53, and week 77, and plasma at baseline, week 53, and week 79. There was a significant reduction in p-tau (p-tau 181), measured via plasma sample and CSF, in the lecanemab group (-0.6 and -15.9 in plasma and CSF, respectively) compared to increases the placebo group (0.2 and 12.9 in plasma and CSF, respectively) (p<0.001 in both analyses).3® Neurogranin was examined by CSF at baseline, week 53, and week 77. There was a significant decline in neurogranin the lecanemab group compared to an increase in the placebo group (-71.9 vs 18.9, p<0.001).** Neurofilament light chain (NfL) was examined by CSF at baseline, week 53, and week 77, and plasma at baseline, week 53, and week 79. There was no significant treatment effect on both of the NfL measures.** All biomarker outcomes are provided in the Supplement Table D3.4. Additional Harms ARIA was assessed by MRI at baseline and five additional time points across the study. In terms of radiographic severity, ARIA was measured on a 3-point scale from mild-severe. Symptomatic ARIA- E was assessed by radiographic severity and self-reported symptoms (both participant and assessor were blinded to treatment group). If a patient experienced asymptomatic, radiographically mild ARIA-E, they continued the drug with additional MRIs at 30, 60, and 90 days after ARIA was identified. If ARIA-E worsened or became symptomatic, the study drug would be temporarily discontinued. If symptomatic or radiographically moderate or severe ARIA-E occurred, the study drug would be stopped, and MRI was conducted every 30 days until it had resolved. Upon resolution, participants resumed the study drug. If occurred more than twice, the participant would ©Institute for Clinical and Economic Review, 2023 Page D17 Alzheimer's Disease - Final Evidence Report Return to Table of Contents be discontinued from the study. If severe ARIA-E was associated with serious adverse event, the study drug would be stopped permanently. Safety was also monitored in an unblinded manner by an independent monitoring committee. As noted in the main report, participants in the lecanemab group reported more infusion-related reactions compared to the placebo group (26.4% vs 7.4%).3% These reactions mostly occurred on the first dose (75%) and were mostly mild-moderate in severity (96%), with 0.8% participants in the lecanemab group reporting a severe infusion-related reaction.* See Table D3.11. Of those who had an infusion-related reaction, approximately 40% received preventative medications (i.e., anti- inflammatory drugs, antihistamines, or glucocorticoids) and there was a 35% recurrence rate which was similar regardless of whether the participant took the preventive medication or not. CLARITY AD Open-label Extension Phase At the time of this report, the open-label extension (OLE) phase is still ongoing. In this phase, clinical and safety assessments were conducted every six months. The safety MRI schedule followed the same schedule as the core phase (e.g., taken at week 9, 13, and 6 months) for the first six months and then every six months after. The management of ARIA was the same across the core and extension phase. Phase II: G000-201 G000-201 was a Phase lIb multi-center, placebo-controlled, dose-ranging, randomized control trial that aimed to identify the most effective dose of lecanemab. Six doses of lecanemab (2.5 mg/kg biweekly, 5 mg/kg monthly, 5 mg/kg biweekly, 10 mg/kg monthly, 10 mg/kg biweekly) were tested versus placebo for 18 months.2%? This trial utilized a Bayesian response-adaptive randomization whereby there was a fixed period of randomization for the first 196 participants (N=28 in each lecanemab arm, and N=56 in placebo) and then the response-adaptive randomization was implemented to minimize the overall sample size and study duration. See section on "Bayesian adaptive randomization" below for additional description of this randomization procedure. The primary outcome was change in ADCOMS at 12 months. This trial also included an extension phase of the 10 mg/kg biweekly lecanemab dose only. Participants had an off-treatment period and then re-initiated treatment for up to 60 months. The primary outcome for the extension phase was to examine adverse events and change from baseline in biomarkers at 60 months. Inclusion/exclusion criteria and baseline characteristics are described in Table D3.1 in this Supplement. Participants were included in the Phase llb study if they were aged between 50 and 90 years with MCI due to Alzheimer's Disease or mild Alzheimer's Disease dementia, had objective impairment in episodic memory and MMSE score 22-30, had a positive amyloid loid as determined by PET or CSF [AB(1-42)], had a BMI greater than 17 and less than 35 at screening, and had a study partner. Participants were not able to use therapies such as: immunoglobulin therapy (6 months), biologic ©Institute for Clinical and Economic Review, 2023 Page D18 Alzheimer's Disease - Final Evidence Report Return to Table of Contents drugs (6 months), and anticoagulants (e.g., warfarin, dabigatran for 7 days or 5 half-lives). Other exclusion criteria included any neurological condition that may be contributing to cognitive impairment, any psychiatric diagnosis or symptoms, contraindications to MRI, evidence of other clinically significant brain lesions, or being on any Alzheimer's Disease medication for less than 12 weeks. There was a total of 854 participants who took part in the G00-201 trial {lecanemab n=609 and placebo n=245). Participants had a median age of 71.7 years (range from 50-90) with 64% described as MCl due to Alzheimer's disease. A total of 54.7% of participants were currently on Alzheimer's disease medication and 71.8% were ApoE €4 carriers. Participants reported a mean ADCOMS of 0.38 (SD=0.17), mean MMSE of 25.7 (SD=2.4), mean CDR-SB of 2.95 (SD=1.42), and mean ADAS- Cogl4 of 22.5 (SD=7.5).102 Bayesian Adaptive Randomization The response-adaptive randomization aimed to find the simplest (e.g., monthly vs. biweekly) and most effective (e.g., 10 mg, 5 mg, etc.) dose (= 90% of the maximum treatment effect) and allocate more participants to this dose at each interim analysis, which was conducted every 50 participants.1®® The trial also monitored for futility (<5% posterior probability that most likely effective dose is superior to placebo by 25% for the first three interim analyses and increased to <7.5% once they had reached 350 participants) and success {from 350 participants: >95% posterior probability that effective dose was better than placebo, or at the end of the trial: lecanemab effect exceeds placebo rate by 225%) during the interim analyses. The modeling had two components used to estimate the primary endpoint at week 52 for the adaptive randomization. First, the investigators used a dose-frequency model that examined mean change from baseline. This Bayesian model used weak prior distributions for mean response and allowed borrowing/smoothing across neighboring doses and frequencies to provide a superior estimate of each treatment arm's effect. Second, the investigators used a longitudinal model that examined correlations between early ADCOMS scores and the 52-week outcome. To do this, the investigators used priors selected from historical data and a Bayesian imputation within Markov chain Monte Carlo to impute the 52-week data and update the dose-frequency response model at each interim analysis.'%® As a result, the size of the individual dose groups differed depending on the interim analyses. Notable protocol changes occurred during this Phase Il trial, during the randomization period of participants 300-350. The regulatory authority requested that ApoE £4 carriers (approximately 70% of the overall population) no longer be administered the 10 mg/kg biweekly dose of lecanemab going forward, due to increased risk of ARIA. As a result, 25 participants who were ApoE €4 carriers were discontinued as they were in the trial for less than six months, compared to 46 ApoE €4 carriers who were already on this dose for longer than six months and were able to continue. 12 ©Institute for Clinical and Economic Review, 2023 Page D19 Alzheimer's Disease - Final Evidence Report Return to Table of Contents The regulatory change may have enriched the intervention arm with participants in the lecanemab group expected to perform better. Results In this supplement, we primarily focus on reporting the results from the lecanemab 10 mg/kg biweekly dose versus placebo, as this is the dose that was used in the Phase Ill CLARITY AD trial. Where that data is not available, we provide pooled data from the 10 mg/kg biweekly and monthly lecanemab arms versus placebo. Cognitive Outcomes The primary outcome in G000-201 trial was the change from baseline in ADCOMS score at 12 months. Bayesian analysis was used to examine the primary outcome (ADCOMS) at 12 months. The 10 mg/kg biweekly dose was identified as the most effective dose. The 10 mg/kg biweekly group had a 64% probability of being more effective than placebo with 25% less decline on ADCOMS at 12 months. This result did not meet the pre-specified 80% probability threshold and thus the primary outcome was not met. The 10 mg/kg biweekly group had a 76% probability of being more effective than placebo with 25% less decline on ADCOMS at 18 months (secondary outcome), which also did not meet the pre-specified 80% threshold. However, the 10 mg/kg biweekly group had a 97.6% and 97.7% probability of being superior to placebo at any magnitude on the ADCOMS at 12 and 18 months, respectively (Table D3.7-8.). Subgroup analyses were conducted for those who were ApoE g4 carriers and non-carriers. In ApoE g4 carriers, the 10 mg/kg biweekly group had a 93.6% probability of being more effective than placebo with 25% less decline on ADCOMS at 18 months and a 99.2% probability of being superior to placebo at any magnitude on the ADCOMS at 18 months. In contrast, non-carriers had 29% probability of being more effective than placebo with 25% less decline on ADCOMS at 18 months and a 63% probability of being superior to placebo at any magnitude on the ADCOMS at 18 months (Table D3.13). Thus, the analyses suggest a potential differential treatment effect for carriers versus non-carriers, with ApoE &4 carriers experiencing a greater treatment effect. However, this subgroup analysis should be treated as exploratory as the ApoE €4 carrier group consisted of 45 participants (non-carrier n=107) due to the regulatory changes that prevented ApoE €4 carriers being randomized to the highest dose. Bayesian analysis was also used to examine other secondary outcomes (CDR-SB and ADAS-Cogl14 at 18 months). At 18 months, the 10 mg/kg biweekly group had a 96.4% probability of being superior to placebo at any magnitude on the CDR-SB and a 98.8% probability of being superior to placebo at any magnitude on the ADAS-Cogl4 (Table D3.14-15). ©Institute for Clinical and Economic Review, 2023 Page D20 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Alongside Bayesian analyses, frequentist mixed effect model with repeated measures (MMRM) were conducted for secondary outcomes comparing placebo to the lecanemab arms at 18 months. The MMRM included the following variables in the analysis: treatment group, visit, clinical subgroup (MCI due to AD, Mild AD dementia), the presence or absence of ongoing AD treatment at baseline, ApoE €4 status, world region, and treatment group by visit interaction as factors, and baseline value of the cognitive outcome as covariate. The analysis did not correct for multiplicity. Additional sensitivity MMRM analyses were conducted, such as including the interaction between ApoE4-by- treatment-by-visit interaction which were important as randomization was broken and would have accounted for baseline risk and rate of change differences. In a manuscript that compared the Bayesian analysis to the MMRM analysis in the G000-201 trial, the authors reported consistency of the cognitive effects across the two statistical methodologies and that Bayesian analysis was useful in accommodating missing data from the protocol change imposed by the regulatory authority.1%4 Participants who received lecanemab 10 mg/kg biweekly had less cognitive decline in ADCOMS, as represented by a smaller increase in the score, at 12 months compared to placebo, mean difference versus placebo: -0.05; 95% Cl: -0.8 to -0.01; p=0.027.1%2 These results were consistent at 18 months with those in the lecanemab group showing 29.7% reduction in cognitive decline as compared to placebo (37.4% in the MMRM analysis that included ApoE4 status and visit interaction)%4, mean difference versus placebo: -0.06; 95% Cl: -0.10 to -0.01; p=0.034 (Tables D2.1. and D3.4-5). There was no significant treatment benefit of lecanemab 10 mg/kg biweekly on CDR-SB at 18 months, mean difference versus placebo of -0.396; 95% Cl: -0.82 to 0.03; p=0.13 (Table D2.1).192 There was a 26.5% reduction in cognitive decline on CDR-SB in lecanemab 10 mg/kg biweekly at 18 months compared to placebo (32.1% in the MMRM analysis that included ApoE4 status and visit interaction).1% There was a significant treatment benefit of lecanemab 10 mg/kg biweekly on ADAS-Cogl4 at 18 months for lecanemab as compared to placebo, with a mean difference of -2.31; 95% Cl: -3.91 to -0.72; p=0.017.1%2 There was a 47.2% reduction in cognitive decline on ADAS-Cog14 in lecanemab 10 mg/kg biweekly at 18 months compared to placebo (55.9% in the MMRM analysis that included ApoE4 status and visit interaction) (Table D3.4-5).1%4 Subgroup analyses reported data on the treatment benefit of lecanemab 10 mg/kg biweekly on cognitive outcomes {CDR-SB, ADCOMS, and ADAS-Cog14) for those with MCI due to AD or mild AD. The subgroup analyses provided conflicting results of efficacy (see Table D3.13.-15 for full details).1%? A subgroup analysis was also conducted on the ADCOMS outcome, specifically, and reported no difference in efficacy for Asian participants as compared to the full sample.'%® These analyses were conducted in small groups, likely underpowered, and thus should be considered exploratory. Beta-amyloid ©Institute for Clinical and Economic Review, 2023 Page D21 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Decreases in beta-amyloid, as measured by Florbetapir PET SUVr, were significantly greater in the lecanemab 10 mg/kg biweekly group compared to placebo at 12 and 18 months (12 months: -0.26 vs placebo: -0.01, p<0.001; 18 months: -0.31 vs 0.004, p<0.001.19%1% The investigators reported that 65% of those in the lecanemab 10 mg/kg biweekly group at 12 months and 81% at 18 months were amyloid negative by visual read.1%6:297 Decreases in plasma AB42/40 ratio were significantly greater in lecanemab 10 mg/kg biweekly group compared to placebo at 12 and 18 months {p<0.01).1% There was also a significant increase in AB (1-42) in the pooled 10 mg/kg biweekly and monthly lecanemab group, compared to placebo (Table D3.4).1°2 The investigators suggested that this increase may reflect changes in amyloid aggregation or normalization of amyloid levels. However, the analyses for these two outcomes were conducted in a subset of participants and thus it is not possible to infer conclusions about these data. Other Biomarker Outcomes In the pooled 10 mg/kg biweekly and monthly group, there was a significantly larger decrease in p- tau as compared to placebo at 18 months, mean difference: -12.3; p=0.013 (Table D3.4).102 However, there were no significant differences in the change in t-tau, neurogranin nor in neurofilament light chain in the pooled group as compared to placebo.1921% These analyses were also conducted in a subset of participants and, as previously noted, it is not possible to infer conclusions from these data. Brain volume was assessed using MRI scans at screening and months 6, 12, and 18. Volumetric MRI data was available for 72 participants in the lecanemab 10 mg/kg biweekly group and 162 participants in the placebo group. There was a greater decrease in whole brain volume in the lecanemab 10 mg/kg biweekly group compared to the placebo group at 18 months (mean difference: -8118.34 mm3; 95% Cl: -10538.26 to -5698.42, p=0.001) (Table D3.4). There was a greater increase ventricular volume in the lecanemab 10 mg/kg biweekly group compared to the placebo group at 18 months (mean difference: 2317.96 mm?3; 95% Cl: 1678.88 to 2957.03, p=0.001). There was no significant difference between the groups in change in hippocampus volume at 18 months (mean difference: -19.44 mm?3; 95% Cl: -46.77 to 7.88; p=0.24).102 ©Institute for Clinical and Economic Review, 2023 Page D22 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Table D2.1. Key Trial Results Measure Lecanemab (G000-201) Intervention (10 mg/kg Q2W) | Placebo Timepoint 12 months Baseline (SD) 3.0(1.4) 2.9(1.5) N 84 161 CDR-SB Mean change (SE) 1.10 (0.21) 1.5 (0.16) Difference in mean change (95% -0.396 (-0.82 to 0.03) REF Cl) Timepoint 12 months Baseline (SD) 0.37 (0.15) 0.37(0.17) N 93 187 Mean change (SE) 0.09 (0.02) 0.13 (0.01) Difference in mean change (95% -0.05 (-0.08 to -0.01) REF ADCOMS Cl) Timepoint 18 months N 79 160 Mean change (SE) 0.14 (0.02) 0.19 Difference in mean change (95% -0.06 (-0.10 to -0.01) REF cl) Note. Ranges for cognitive scores at baseline were not reported. ADCOMS: AD composite score, CDR-SB: Clinical Dementia Rating scale Sum of Boxes, Cl: confidence interval, mg: milligram, kg: kilogram, Q2W: every two weeks, SD: standard deviation, SE: standard error Harms Total adverse events, serious events, and deaths were similar across the lecanemab 10 mg/kg biweekly and placebo groups. The rate of total discontinuation was greater in the lecanemab 10 mg/kg biweekly group compared to placebo (14.9% vs 6.1%).1%%2 There was a higher prevalence of discontinuation due to adverse events in the lecanemab 10 mg/kg biweekly group compared to placebo (13.7% vs 6.1%), with most of the adverse events in the lecanemab group relating to ARIA-E (Supplement Table D2.2). Several participants discontinued due to "other" reasons which the investigators described as due to the regulatory change (ApoE €4 carriers in the highest dose and were on treatment for less than six months were discontinued from treatment), subject moving out of the area, or the loss of a study partner. The lecanemab 10 mg/kg biweekly group were also more likely to experience infusion-related reaction compared to placebo (19.9% vs 3.3%). From this, there were 2.5% of participants in lecanemab 10 mg/kg biweekly who discontinued due to infusion- related reactions, compared to 0.80% in the placebo group.1® ARIA was assessed by MRI at baseline and seven additional time points across the study. In the core phase, if a patient experienced ARIA-E they would be discontinued from treatment. During the open-label extension phase, if ARIA-E was asymptomatic or radiographically mild-moderate then the participant could continue with monitoring. If symptomatic or radiographically severe, then treatment would be paused until ARIA-E resolved and then treatment could be resumed. If ©Institute for Clinical and Economic Review, 2023 Page D23 Alzheimer's Disease - Final Evidence Report Return to Table of Contents occurring more than twice, the participant would have been discontinued from the treatment. During the open-label extension phase, the protocol was slightly different to the core phase in that if ARIA-E was mild-moderate or asymptomatic then participants could continue the intervention with monitoring. But, if ARIA-E was severe or symptomatic, then the intervention was temporarily discontinued for these participants until ARIA was resolved and then they restarted on the same dose. A total of 9.9% participants in the lecanemab 10 mg/kg biweekly group experienced ARIA-E, compared to 0.8% in the placebo group. Due to the protocol change that prevented ApoE g4 carriers being randomized to the highest dose, only 30% of participants in the 10 mg/kg biweekly were ApoE €4 carriers and thus it is not possible to examine the true prevalence of ARIA-E in ApoE €4 carriers in this dose. To provide some estimation as to the prevalence of ARIA-E in ApoE €4 carriers, we reviewed ARIA in the second highest dose. In the 10 mg/kg monthly dose, 9.9% of participants experienced ARIA-E and 92% of those participants were ApoE €4 carriers. All ARIA-E events in the lower lecanemab doses occurred in ApoE €4 carriers. Additionally, there were 6.8% in the lecanemab 10 mg/kg biweekly group and 11.1% in the 10 mg/kg monthly dose who experienced ARIA-H, compared to 5.3% in the placebo group. There were 4.3% and 5.1% of participants in the lecanemab 10 mg/kg biweekly and 10 mg/kg monthly groups who experienced ARIA-E with ARIA-H, compared to 0.4% in placebo (Supplement Table D2.2). As noted in the main report, symptomatic ARIA-E was of particular interest. Symptomatic ARIA-E was reported to generally consist of headache, visual disturbances, or confusion.%? Symptomatic ARIA-E occurred in more participants in the lecanemab 10 mg/kg biweekly group (1.2%) and in 11% of all lecanemab doses, compared to no cases in the placebo group. There were no cases of symptomatic ARIA-H in the lecanemab or placebo groups. Per protocol requirements, all those who experienced symptomatic ARIA-E were discontinued from the treatment. The investigators provided no further information on the prevalence of symptomatic ARIA-E in ApoE 4 carriers and thus we were not able to examine this. ©Institute for Clinical and Economic Review, 2023 Page D24 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Table D2.2. Adverse Events Adverse event Lecanemab (G000-201) Lecanemab | Lecanemab | Lecanemab Lecanemab Lecanemab 25 Qr;'illkg 5 anf'{vkg 5 gzg'{vkg 10 mg/mg QAW 10 anlgVCkg Placebo Bentoeg | @os | et | meo | PB4 | 0o | e naversevent, |92 ysiom |92 |y | BS | n/N (%) Due o Adverse | 7/52035%) | 4517 | 12 |apsssen |21 | 157208 Event, n/N (%) :;'""' '?;)'A'E' 1/52 (1.9%) | 1/51(2.0%) ?3/22%) 25/253(9.9%) | 16/161 (9.9%) (Zgi':/':') Z':s::':;fii%) 1/52 (1.9%) | 0/51(0) (11/%2%) 1/253 (0.4%) 2/253 (0.8%) | 0/245 (0) :;';?;;A'H' 2/52 (3.8%) (71/;17%) (117;?%) 28/253 (11.1%) | 11/161 (6.8%) (153{,'29:)5 ARIA: Amyloid-related imaging abnormalities, mg/kg: milligram per kilogram, N: number of participants, Q2W: biweekly (once every two weeks), Q4W: once every month *Incorrectly reported in Swanson et al. (2021) as 39/161. G000-201 Open-label Extension Phase The OLE phase included 130 participants from the core study (n=42 in prior placebo group, n=37 in prior 10mg/kg biweekly group, and the remaining 101 participants were in the other lecanemab doses).1° Baseline characteristics for the extension phase are reported in Supplement Table D3.3. Of note, at the OLE baseline, the majority of those who had received lecanemab 10 mg/kg biweekly during the core phase and continued onto the OLE phase were ApoE4 non-carriers (91.9%). While it is expected that there would be fewer ApoE4 carriers in this group due to the regulatory changes that prevented ApoE4 carriers being randomized to this dose early in the trial, only 3 out of 46 ApoE4 carriers in this dose continued onto the OLE extension. It is unclear the reason behind participants not continuing onto the extension phase, but may relate to tolerability, adverse events, such as ARIA, or lack of efficacy. Connected to this final reason, participants who received lecanemab 10 mg/kg biweekly during the core phase and continued onto the extension phase had lower levels of amyloid (as measured by centiloids) at extension phase baseline compared to placebo and those who received any of the other lecanemab doses. However, the standard deviation of the mean centiloids was very large suggesting that there was considerable heterogeneity in the clearance of amyloid in this group. ©lnstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page D25 Return to Table of Contents All participants had a gap period off-treatment, with a mean duration off the study drug of 23.7 months (range: 9.2 to 52.5).2% During the gap period, cognitive ability, as measured by ADCOMS, worsened in all groups with a similar rate of progression across both those who did or did not receive lecanemab in the core phase, and amyloid, AB42/40, and p-tau increased in those treated with lecanemab during the core phase (Table D3.6). During the OLE period, participants received lecanemab 10 mg/kg biweekly, including ApoE £4 carriers, for up to 60 months. Clinical assessments were administered every six months. Once redosing begin during the OLE phase, amyloid, AB42/40, and p-tau decreased in the two groups (those treated with lecanemab or placebo during the core phase).1%1% At month 12, by visual read, 83% (n=10) of those in the core placebo treated group had converted to amyloid negative status.'® In terms of clinical outcomes, by 18 months, participants in both groups showed a similar rate of clinical decline.?'® See Table D3.4. Itis to be noted that these analyses were based upon very small numbers and thus caution should be taken in interpreting these values. In terms of harms, safety assessments were monitored slightly differently to the core phase. Hematology, blood chemistry, and urinalysis were assessed at baseline, weeks 3, 7, 13, 19, 27, and every 6 months thereafter. ARIA was also monitored and managed slightly differently to the core phase. In this phase, safety MRIs were taken at baseline, week 9, 13, 27, and every 6 months thereafter. In this phase, if ARIA-E was mild-moderate or asymptomatic then participants could continue with treatment. If severe or symptomatic, dosing was paused until ARIA resolved and then participants could restart treatment. ARIA rates were similar to the core phase with 7.8% of ARIA-E in lecanemab core phase treated participants and 8.9% in placebo core phase treated participants, all of whom were ApoE €4 carriers, with around 2% having symptomatic ARIA.106:111 Similar to the core phase, most ARIA-E occurred within first 3 months and resolved within 12 weeks. At the time of this report, the investigators reported that 6/14 mild-moderate cases continued treatment through ARIA. A total of 20.6% had infusion-related reactions during the OLE phase that were mostly mild-moderate.2%° Phase I: Subcutaneous Dose An open-label parallel-group trial including 59 healthy participants was conducted to examine the bioavailability, pharmacokinetics, safety, and immunogenicity of a single 700 mg subcutaneous {SC) dose of lecanemab. There were 30 participants who received IV infusion and 29 received SC injection. In terms of safety, adverse events were comparable across the two dosage forms and 20.7% in the SC dose group reported injection-site reactions, compared to 33.3% reporting infusion- related reactions in the IV dose group (Table D3.12).11? Investigators concluded that SC dosing appears to be a potentially feasible option to progress into Phase Il trials. ©Institute for Clinical and Economic Review, 2023 Page D26 Alzheimer's Disease - Final Evidence Report Return to Table of Contents S]U9IUO) JO 9|gE] 0} uIMay Hoday 23uapIAg |eul{ - aseasiq S, JawIdyz|y £Tq a8ed €20C 'M3IA3Y d1WOU0dT pue |ediul]) 04 SINNISUIQ uedpiued 1Uasuod ay1 ul saunpasoud Apnis pawojul Jo awil ay} YHM 219JJ31ul P|NOI 1eY] 1e 'siedA pg (=>) 01 |enba 'sajpoqaue (suoisn|ap 40 'uojssaidap JO uey1 ssa| pue Qs (=<) 01 |euojdouow syjuow Jofew 'suoneupnijey |enba Jo uey) Ja1ea48 pade J1WaIsAS 8T 1€ 1av-DN 'a/dwexa) swoldwAs | sjuedpnaed ajewsy Jo 3jeN 'suiqoj3ounwiw) -2QAV pue 'SINODQAV | Jo sisouBelp auaieiydAsd Auy A3ojoyied piojAwe | Saideiayl pauqiyoid 'vy180D)-svav '13d Sujuaauds Jo | uleiq Joy JaxJeWOlq BAILSOd plojAwe uj aduey) Syjuow T Ulyum sainzjas (1A AI-SIANM) CILEE :(2102) Asepuodas Jo '@jouis '(v|1) syoene | || (3jeasqns) Alows|y [e2180] aJe s123(qns aAleu [syluow G ] GS | d1wayds| JudISURI] JO AJOISIH | -Al 91BIS AIOWDA JB|SYIIM | -Juswiieal] 'Bujaseq odd -¥a) ui adueyd 's3y aseas|p 9y} u] ueaw paisnfpe-ade 0} Joud syaam (eseyd (uoisuaxa) Alewnd | sJawayzly s,auedpiued ay MO[2q UOIIBIASD pJEpuURlS | 2T => .10} ludwiea) uo|su9IXa) [syruow Aq pasneo jey} puoAaq pue T 1se9| 1e Aq pa3eoipul aV JUBHWOdUOD MZO aseyd 00:120Z AVL) 8T] 8S-4aD ul | SAoqe juawuiedwi aAlHuSod se AJowauw Jjposida Jo 9sop a|qels Sw o1 91 10 |e 19 uosuems sujaseq woJ) a8uey) | 03 8unnquiuod aq Aew eyl ul Juawjedwi aABlqoO saidessyl MO | +2d I | 'no8sjeipjesiul) :(Apnis au02) Asewd uollpuod [eai8ojoinau Auy Apnisaio) punoisjoeq Sw ot 901 aseyd av ALNv1D suojjeaipaw |euoiresiisaaul o< || INT Y-SIAIM - 'saniua [7 =< 2J02s ISNIA - |estways mau aniysod AH - 03J02103S HYAD - | 'gewsduedd| 'Adesayy [syoom sainpasoud 30dV T 1sed| je sassassod uigojSounwuwij 912/96] 14D '13d Apnis yum auspialu| 'ejlpuswap yum pasouselp 'spiojAwe-jjuy odd ui 88uey) :Aiepuodas pIno3 1ey3 swoldwAs dAIle[24 99480p 15.1) se | SoidelayL pauqIiyoid MPO 8w [syo9m 912] Jo sisou8eip suieiyoAsd yons SJ0108.} ySI [RUOIHppR OT + 8w s g1 194 13d u1 88ueyd '5HIvd Jo Aio3siy uaun) - dABY ISNW $9 0} GG - Jedpun MZ0 8w 'a0 'lll | AoBsjenpjeaulp ul 98uey) :Alewnd | Syae1Ie d1waydst Jo A0ISIH - sJeaA g 031 G5 pasdy - eJay] punoidydeg | OT + 8w § gD aseyd SH-€ AVIHY qewaued'a [saujodswiy] sjuajjed papn|ox3 sjuajzed papnpou| Adesayy suy udisaq lent saw021nQ Ad) ¥ : punoidydeg jJuawieadl] Apmis ' usisag Apmis '1°€a 3|qeL s3|qeL Ju3pIng "€d S]UaIUO) JO 9|gE] 0} uImay 8¢Q aded 1oday a3uapiag |eul{ - asessiq S, JawWidyz|y €707 'M3IASY J1WOUODT pue [edjul|] 10} SINIISUIQ [syuow 8T pue zT] sJ9yJewolq Alojesojdxa pue '[syluow 8T pue ZT] [HIN ela SWN|OA ujelg-3joym pue "Jejnaliusa 'ledwedoddiy '13d plojAwe uelq 'pT802-SvVAY 'dS-4ad '[syauow 8T] SINODAaV ul duljaseq wouy 98uey) :Auepuodas [syruow 7] SINODQV uo aul|aseq wouy 38uey) :Azewid elJUSWp € 31edipu] pIN02 1eY) SUOIST| JuedJIusIS Ajjeaiuifs J2Y1o Jo 93U3PIAT - Suluuess YAl 01 SsuonedIpulesIuo) - 8294005 50O - saJnpasoud Apnis yum auapau| pIno3 1ey3 swoldwAs Jo sisouSeip oieydAsd Auy - JeaA T 1sed 9Y1 U] 59uNzI9S JO 'D0.IS 011y J1WaYds| Jualsue.d ) JO 3041S B peY dABH - jusuneduw 9AIu809 s,199[gns 9y3 jo asned 8uinquuod e 3q ySiw 1ey) (aseasig s, Jowayz|y ueyl Jaylo) uoppuo3 |esi8ojolnau Jo |eaipaw Auy - suojled|psw Qv panoidde 10 950p 9|qe1s 40 SAIEN - ¢ IS - AT AIFSWM) Adowaw J|posida uj usuedw| - 480 10 134 :zi-TgV ein aamisod piojAwe pawuo) - epuswap av piiw Jo DN 01 3np gv - (428uo0| S1 J9ASYIIYM 'SBAl| -Jjey g 10 sAep / 1o} uesiedigep 'uliepiem '89) syue|ndeodnuy pue '(syjuow 9) s8nup 2180|019 '(sywuow g) Adesayy ulingojdounuwi| S[delayL payqIyoid SAlEU JUBW3IEaJ] 10 suoljeaipaw av panoidde JO asop 3|qels saidessyl punoisyeg odd M20 8w o1 901 MYO 8w o1 901 MZO 8w 0g 911 MYO 8w 0g 91 MZO 8w gz 2d 'ad W "qll aseyd 2011202 |e 1@ 'uosuems T0¢ -0009 Il ®seyd 9SBISIp SJoWIdYZ|Y uey} Jaylo sisousdelp BIJUSWAP B 31edipul pIno3 eyl Suiusauos 1e [YN uielq uo suo|sd] Juediusis Ajjesiulpp Jay3o Jo 33uapiag (s1ouueds YAl Ul 3sn Jo} 9jes se panoidde asoyy ueyl 19430 SDIAJP JelpJed pue [Inys ui 3jdwexo) syuejdwi |ersw J119uSewo.ludy 'Jo1e||11q1)op/iexewaded aejpied Suipnpui 'Bujuueds I 03 SUOIIBJIpUIRIIUO) 8ujuaauds 1e g< 21035 (Sao) 9|eas uojssasdaq aulelan GE> LI<INE 0€> TZ< ISWIN J)Jewolq aAIUSOd - BEIVETG) Apnis 3107 pa1a|dwo) - 35eyd UOISUBIXT Suiusauos 1e gg (>) ueyy Ss3| pue £T(<) ueyy 1931e2.3 (1ng) xapuj ssew Apog auljaseq pue ujuaalds e Og> pue suljaseg pue 8ujusa1os je Zz3 94035 (ISININ) uoneujwexa 931Ee3S |RIUAW JUIN S]UaIUO) JO 9|gE] 0} uImay 6¢d a3ed 1oday a3uapiag |eul{ - asessiq S, JawWidyz|y €707 'M3IASY J1WOUODT pue [edjul|] 10} SINIISUIQ Aj@1e1ndde 53591 Sulwiopad juailed juaaaad pjnod eyl SUOJ}HPUOD |EJIPAW JBYIO - Dl3 eia |eAsD1ul 21D/10 paduojoud - avy ueyz Jaylo sisoudelp S]UaIUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S JawIdyz|y ¥1802-Syav 'as-4ad 'SINODQ@V ul 88uey) "[sysuow o9 03 dn[ awi} Jano siuedpiued aasod plojAwe Jo 98ejuadiad 'aseyd UOoISUIIXD JO Aulj3seq 03 Apnis 3102 Jo pud wouj plojAwe ujeuq ul 98ueyd '[syuow ¥Z pue '¢T '9 '€l S|9A9] plojAwe uleuq uo aujjaseq wWouy a8uey) :Alepuodasg [syjuow 8/ 01 dn] y|yy Jo SJUBLUSSASSE |YIN pue elep Ayajes :Atewld Sujuueds IHINl 01 SuoedIpUIRIIUO) - 82 9400s 50O - sainpaosoud Apnis yum alsapaul pInod jey3 swoldwAs uo sisouSeip aieiydAsd Auy - JedA T 1sed 9Y3 Ul S2UNZISS JOo '9)0J1S oe11y 21WIYIS| Judlsued JO 3)0.43S B pey dAeH - juduLeduw) 9Anu3oa s,199[gns 3yl Jo asned Sulnguuod e 9q 31w 1ey) (aseasig s, JowIdyz|y ueys Jayio) uollpuod [ea18ojoinau Jo |eaipaw Auy - suoljesipaw 71y panoidde JO 9S0p 3|qe3S 10 dAIRN - CC IS - AT AI-SIIMW) Adowaw J1posida uj Juawuedw| - 450 10 134 :zi-TgV ein aasod piojAwe pawauo) - ej3uswWap 771V pliw Jo QA 03 @np @V - uay) aulsseq 310 1e syuengeod|jue uo J1 'uojlpuod |eaipaw aya jJo uonn|osal 10 uolezijiqels [13un papuadsns Ajuesodway 9q |I!m 8nup Apnas s8nup onAjoquouayy uo JI '(syjuow 9) s8nup 2180|019 '(syruow g) Adesayy ulingoj8ounwiw| SSIdEIa|L paNqIyoid Jeapun saidesayl punoJspeg syluow 09 10} Juswieas Suneu -91 210j3q (syjuow 1z 98esane 'syuow 65-6) polad JusWIR3N-JO MTO Sw ot 901 370 '2d 'aa 'il aseyd 0€Q 98ed €70¢ 'M31A9Y d1WOou0d3] pue |edjul[) JO} 91NUISUIQ Aj31e4Nn228 $3591 Sulwiopad juaiied juaaaad pjnod eyl SUOJHpUOD |EJIPBW JBY1Q - D3 Bl |eAs1ul 510/10 pa8uojoud - -Buiuaauas ay uey1 Jaylo sisouselp 370 240439 S}3am elIUSWp kB 31e3|pul ¥ 1sea| 1e 10} 3|qe1S pIno3 1ey) suoisa| Juediusis pue paziwnido sniels '[syauow g9 01 dn] | Ajjediuld J2Y10 JO SDUIPIAT - uone|ndeospue aog'sjeyeanulp I10 102 -0009 || 3seyd SJUaIUO) JO 9|geL 03 UIN1ay 1oday 27UapIAT [euUl{ - seasI( S JaWIsYyzZ|y T€Q 98ed €207 'M3IASY J1WOUODT pue [edjul|] 10} SINIISUIQ uollp3 Yino4 - a|eas AJOW A J3|SYI9M AI-SIAIM "Snoaueinagns :JS 'syasm Jnoy AIaas :pPD 'Syoam oml Alans :pZD 'Aydes3ow o] uoissiwg uousod :13d 'pajjosiuod-ogqadeld :Dd 'S ausodwo) 9AINUS0D JaWIBYZ|y [ea1Ul[IR.d :§IIVd 'UoISUaIXd [aqel-uado ;370 'panodad Jou YN 'BuiSew adueuosad d11pusew (YN 'wesSo|ry Jad wead)w 8y /3w "uawiedw] 3AINUS0D pliW (DAl "4SIURDINW DA 'WesS|jjIw DA 'UOIFBUILUEXD 1E3S [BIUSW JUIW :ISIAIN 'gBWaueId)| :g)7 'SNOUSARIIUL iA] 'WelS01pied0J13]3 523 'pullg-3|qnop :dq 'saxog Jo wing - Zupey eluawaq [edulD :@S-4ad 'Suney epuawaq [edlul]d :HAD 'X9pu| UolduN4 3AINUS0)D 114D (UOISI3A DIN) Aojudnu| Suia Ajieq Jo saniARRY - Apnis dAleladoo) aseasiq s, JawiRyz)y 11av-1DIN-SOay 'saiijewloude SuiSew pajejaJ-plojAwe 1y|yy 'Wal-yT 9|easqns 9AI}IUS0) - B]e3S JUBWISSASSY Bseas| S aWIRYz|Y (yT80D-SyqaV '94005 ausodwo) Qv :SINOJAV 'OSessip S, JaW|ayz|y :qy '4olgiyul aselalsauljoydjAiaae :s|Jydy noy Ajp1ewxosdde 12A0 pasnjul Al 83/3w 0T zir (Pensqy Jo asop a|3uis Apnis ERITESET ) ApuaSounwiwi Jaye asop dnou8 JIVY - 2202 'e pue Ala)es '(Md) Al 'udwopge | -|9|jeJed 19 8unoq) asoq soupjodew.eyd 9y} ul asop '°09e] snoauelnaqgns '(va) Aujiqe|ieaeolq 2s8woos | -usdo paxi4 a1n|osqy :sawonng NN 'syuedpiped AyyjeaH UN paxiy aj8uis | '|eseyd | 'e8uis - T aseyd S]U3IUO) JO 9|gE] 0} uImay Joday aouaplIAg |eulq - aseasiq s Jawiayz)y 7€Q 98ed £20¢ 'M3IA3Y d1WOU0D] pue [ed1ul]) JO) SINNISUIQ (9vz) s6LT/1vY UN UN dN $31PIGIOWOD +§ (T'TS) S6LT/L16 UN UN UN SOIUPIQIOWOD +7 (9'91) S6.T/86C UN N UN ANSs00 (T'st) S6LT/1LT dN UN dN sajeqelq | (%) u 'senipiqiowio) (z'91) S6LT/16C UN 4N 4N 9SEasIp 1eay olwayas| (t'09) 56£1/580T 4N 4N 4N ejwapidiuadAH (€79S) S6LT/€66 4N dN 4N uoisuapadAH 7 c (8°L€) TEE (5°8€) 1€€ 4N 4N 4N av PIIN (%)u '28e35 [EOIUID (z'29) v (5°19) 825 (59) vsT (65) 06 (89) 991 av 01 anp P (T°sT) 2eT (8'sT) 9€T ¥N UN UN (9108Az0Wwoy) o/ (%) (5°€S) 89v (T°€s) 9sv iN UN 4N (9108Az0sm18y) /g | N/u '@dAlousnp 30dv (9°89) 009 (689} z6S (t) 691 (0€) o (68) 812 Jsuie) (%)u 'smeas yiody . (%) (S'v) s6.1/08 N N N sjueinSecopuy | U 'UOREJIPINI JSYIO . . %)u 'uonesjpay (5°€S) 89 (0°28) Lvy (¥S) 82T zs) 6L €5) TET Sunuewa Yo/pue ( (cs) (€s) sJoMqIyu] asesaisauljoyd av ejwosuo) d _ 4N 4N 4N 4N 4N ounei/ouedsiH-uoN (%)u 'Aaruya3 (g'zT) 80T (s7zT) LOT dN UN UN oupe/ouedsiy §(o'e) 9z §(ev) L€ UN UN UN Jay10 (ree) Le9 (€9£) 559 4N UN 4N aUYM (%) u "@dey (L) vt (7)ot UN UN UN eg (69T) 8¥T (T°LT) L¥T 4N dN UN uelsy (0°€s) vov (9°18) €vv (85) L€T (zv) v (sv)ott (%) u 'ajewad (82)o1L (6L)v1L (68-09) 2L +(88-15) €2 1(06-€S) 1L (as) ueaiy 'a8v $5/8 1658 8€¢ st 9tz N MPO MZ0 3/3w MPO 834/8w 0932eld 3%/8w QT gqewoaueda 0g328ld 0T gewsaueda 0T gewsaueda swuy Apnis «:(TZ0T "|e 33 Auieiz|) ZzoT ] 9JUDIDJUOD QVLD "4:220T °|& 19 )2AQ uen 20t1Z0Z '|B 10 uosuems 92Inos av ALV 102-0009 lenl sanisiIaYIRIRY) dUIjaseq "z €d d|qel S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y €€Q 98ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ 1Yy817 JuswejyounaN 80°C6 618 4N 4N 4N (as) ueay '|w/8d 'ney-d 00°'ST9 00585 UN N UN (as) uea 'Jw/3d 'ney-y 4$d 00"vEETT 00°£86TT (as) ues 'qw/3d 'op-1dv oOv¥1S 00°L¥S (as) uesiy '|w/8d '¢y-Tdv ¥50'0 Ly0'0 HN UN HN o as) uealA 'lw/3d 'ov/zvgv (£¥T°0) 0O¥ 0 (£¥1°0) 86€°0 (£1°0) LEO (sT0)LEO (sT°0)LE0 (as) uean '2403 SWODAY (679) 601 (99) T1v UN N UN (as) ueaw '103§ 1AY-DIN-SIAY (95°2) LEVT (80°£) svvT (VATAR: Y4 (rovee (eL) 612 (as) uean '21038 yT30D-SYAY g'g o} . O . f . - . - - . . . @M:N £ :mm \Whou ! <0 (ve'T) 22°E 080150 (PETILTE (sT)6C (r1)o€ (eT)6C Y '(as) uesn S gS-4dD o€ ey} o v\ e o\ o o\ - ' o ues 2402 017z (2°2) 9°ce 0£ 03 2Z (T2) 5°St (e2) 09z (r'7)9'se (2 Lst (as) uean S ISININ (€'6T) 69T (z'61) 59T UN UN UN T (%)u (£ 08) 90 (808} ¥69 (¥8) 002 (88) €€T (s8) 0TC 50 '2403§ |eqOIO ¥AD (v'1) s6LT/SC 4N 4N UN $912IPIGJOLOD +§ (£°£) s6LT/6€T N N N S91PIIOWOD + S]U3IUO) JO 9|gE] 0} uImay 1oday 93uapIA7 [eul{ - 3seasiq S JaWIBYZ|Y ¥€Q 98ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ (6°8€) 0L (sov) St (Lsg) ST av PN R - (T'19) OTT (965) TT (€99} LT av o1anp P (90€) S (616) € (982) ZT JapuesuoN (9'ST) 82 (8¢ (6'19) 9T (9108AzowoH) ya /o (%)u 'snaess p3ody (6°€S) L6 0 (s6) ¥ (2108Azoud18H) a/€2 (r'69) S2T (tg)e (r'TL) 0€ Japue) (€°8¥) L8 (98v) 8T (00s) 1T (%) u 'sjewaq (cL)ove (02)69L (z8)81L (as) ueay 'a8y «08T L€ [4 ' N MTD 34/8w 0T qewsueds MTD 85/8w 0T qewaueds| Jolid oqgooe|d 210 Jold suuy Apmis o11220Z '|& 32 dpeadn 22inos 310 T0Z-0009 |eny 310 T0Z-0009 }O solisualdeIRY) BUl|aseq "g"EQ 3|qel *(oge2e|d) pE = U 'gewauedd| pSE =u uonejndod Apnisqns 13d plojAwy# Buissiw/Iay10§ 1w (@8ueu) ueipaNg ot AY@amiq 83/8w 0T p=u 'Ajyiuow B5/3w 0T 68=U 'ogade|d g6=U :Apnisqns |3d PIOjAwy, 1ud2Jad 9 'oiled anjeA ayeldn pazipiepuels (IANS 'UOIIBIASP pJepuels :ds '4a1|l|[iw Jad sweidodid :jw/3d 'ney palejAioydsoyd :neiy-d 'ueas AydesSowoq uojssiws uoualisod :13d 'ogade|d :0gd "4aquinu :N 'pariodal j0u YN 'uswaledw! aAIIUS0d pliw DN 'weado|ry Jad weadijjiw 8y /3w 'weado|y 8y 'saxog JO wng - Suiley erpuswa( [B31UID :dS-Had 'Suney eipuawa (31Ul :ygd 'Oseasiq S, J2WIdYz|y Uo S|eL] [eaIul :avLl) 'pInj [euldsoiqalad 14S) '(uoisian D) AlojuaAau| SulAn Ajleq Jo sailAIDY - Apnis aAl3eId00) aseasiq S JaWIdYz|yY IAav-IDIN-SIAV 'UolIBulWEX 91R)S [elusW Julw SN 'I uialosdodijody :Jody 'Wall-pT 9]eISqNS SAIHUS0Y) - 9IS JUDWISSISSY ASRISI S JaWIdYZ|Y :HT80)-Syay '24005 ausodwo) Qv :SIOIAV 'OSeasip s awiayzly :qv N UN (aT'0)O¥'T (91°0) LET (8T°0)T¥'T (as) uesn IANS #9°6LT010°LT HUETT et roc) e . o3uel (8T E0°SL | 019°9T- (v8vY) T6-LL «(V'LE) 8'¥8 +(0'8€) 0'8, «(S'TY) €06 1Qs) uesw 'spioyIua) 134 vLE 0L'€ N N N (as) ueaAl 'nex-d e 612 HN N UN (gs) uesw 'ureyd ewselq 1Yy817 JuswejyoinanN 8800 880°0 UN uN UN (as) uesin 'ov/zrdv S1U9U0) JO 3|qe] O} uiniay GEQ @3ed Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y £70C 'M3IA9Y JWIOUODT PUE [EJ{UI]D 0} SINHISUIQ (€z0 (£T00 (8200 {2110 I 434 -01/9°0-) S¥0- 01 029°0-) ZOE'0- 3 01 TZ8°0-) 96€°0- 01 £19°0-) 0SZ°0- (12 %s6) 341 99'T 12T (9€T"0) TLT'T MM%" (1z'0) 20T'T (T0)8vzT | (3S)38ueyd ueS §1 84035 85-4@ LS. viL €€C 191 8 6v1T N sawonQ aAlIuso) syjuow 8T ujodawil S6.LT 86€ 8€C ¢St 9v¢ N suljaseq - MyD 3/8w 8%/8w ot pajood I Apeamiq 3y /8w Alyauow 8% /8w swy Apnis 0T gewsaueda qewaueda 0T gewsaueda 0T qewsauesa a'9eCL0T 3IUIIBJU0I UOSUBMS w0 m.v_umcwo'_w \"_uo_wc__o_\,_ 4 flmw.""_"._wh_id .N._< w."onucm._w uod V1 720z 'e 12 yoAq uen S 'ap; TZ0Z 39U3I3}U0D VL) °[e 33 UOSUEMS ',r; TZ0Z } 231n08 av.d "1e 32 Auiag {,,,2Z0T '|e 19 BPPRYQ 5 TZOZ °|B 12 uosuems av ALV 102-0009 el SaWo02INQ qewauedd] "y'ed a|qel MEO0T 2402 Joud Tz=u 'ogade|d Joud 7z=u :Apnisqns | 3d pIojAwy *S9SOP qewaueds| JuaJaIp Sulnedal L6 pue 'Apjeamiqg 83/3w QT qewsueds| g¢ 'ogaae|d Suiaigal Ajsnoinaud gf YUM 370 paJa1us 08T« Jua24ad :94 'o11ea SNjeA yjeldn pazipaepuels (IANS 'UOIIRIASp piepuels :as 4s|lf|iw Jad sweadodid :jw/3d 'ney pajejAioydsoyd :nej-d 'ueas AydesSowo] uojssiwa uoJlisod :13d 'ogade|d :09d "Jaquinu :N 'patiodal j0u YN Quswledw) aAIIUS0D pliw DA 'weaSo|n] Jad weadijjiw 8 /8w 'weidofy 8y 'saxog Jo wns - Sujiey eppusawaq [BIIUID :gS-HaD 'Suiiey eiusawaq (el :¥aD '©seasiq S,49WIayz|y uo s|el ] [eaul :avLld 'piny [euldsoiqalad 145 '(UoisiaA D) Asojuaau| SulAr Ajieq Jo sanIAIY - Apnis aAileladoo)) aseasiq S aWIdyz|y 11av-IDIN-SJAV 'Uoileulwexa 91els |epusw juiw :3SNIA '3 uialoldodijody :30dy 'Wal-HT 9]BISNS SAINUB0) - 3]BIS JUBWISS3SSY aseasid S, JaWisyz|y :#T80)-svay '2402S ausodwo) Ay :SINODAV 'seasip s Jswiayzly :ay Hzo)zt HT0)TT HZo) vt (as) uesiy "UANS 13d ndejaquol4 (67€v) S'vv (60€) 98 o zLL a8uel (Qs) ueaw 'spiojua) (vo)Lo (v'o)90 (€0)90 (as) ueaw '2103S SINOJAY (ovT) T°S€ (8'€T)5°2€ (5°€T) vee (as) ueaw 'a10ds y130)-SYaAvV (se)es (e)os (Te) Ly a8ueu '(gs) ueaw '2108 gS-¥AD (99) L'0¢ (09)z1e (e9)s'TC (as) ueaA '2403s ISNIN - - - .H (82} 89 (£'62) TT (6'2p) 8T . (%) '21095 eqOIS ¥QD (v'vv) 08 (r'1s) 6T (zsv) 61 50 S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y 9€Q 98ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ s3w021nQ projAwy 0€d "sA sulPsp EEN| %LE UN UN UN UN aAIus02 Ul uolINpPayY 434 100°0>d UN UN UN UN anjea-d 210§ EEN| (8zorzTiOT HN UN HN UN (1D %s6) Jia 1av-1DIN-SOAV g'G- g'€- N UN N N (35) @8uey) ueay S 102 9/9 UN UN UN UN N 0€d "SA aulPsp EEN| % HN 413y ¥€0'0=d '%.'6T 827°0=d '%EvT aAIMUS02 Ul uonINpPaY 43y 100°0>d €50°0 434 ¥€0°0 87T°0 anjea-d (Lzo0 {(900°0- (€100 {oto0 434 | -01%£0°0-)SO°0- | ©03TLO0-) 6£0°0- 434 | -0170T'0-) £50'0- | ©31G90°0-) 820'0- (1D %06) Kia ¥1Z°0 ¥9T°0 (v10°) 2ST°0 €610 (zoo)9cT0 (zo'0) 9910 (35) @8uey) uean S 6bL 80/ ST 091 6L 9vT N syjuow 8T syjuow 8T jjodawi | SINODav N UN 6100 413y L20°0 T0T'0 anjea-d (ot0°0- (z100 (o000 N UN | 01090°0-) S€0°0- 434 | - 036£0°0-) 9¥0°0- '450°0-) 620°0- (1D %06) 1A (to0) N N (z10°0) €600 TET0 (zo'0) 5800 (T0°0) ZOTO (35) @8uey) uean S UN HN 857 L8T €6 59T N 4N 4N syjuow ¢zT wodawi - - 0dd "sA auipPap 43y %9¢ N N L10°0=d '% LV 9¢/°0=d %/ aARjuS0 Ul UORANPBY 434 T00°0>d ¥ST'0 43y L10°0 9£2°0 anjea-d (190 (€910 (LT20 «(6£0°T . | 9102 30)- 434 -0} /TT)VYT- | 0106T°C) ¥90°T- A/ 01 0T6°€-) ETEC- 01 G€9°T-) 8/Z°0- (12 %56) s41a S y1200-Svav 85'S 'y (6%5°0) SEL'€ m%mmw (18'0) 885°C (s9°0) vzo'v (35) @8ueyy ueay 1 8€/ £0L gt 85T 6L 9vT N sn-d 'oz cn-d fos 7+ 09d "SA auljdap 43y %LT UN N SZT°0=d '%5'9¢ 65Z°0=d {%/'9T aAR|UB UJ UORINPBY 43y 100°0>d 61T°0 43y SZT'0 SS2°0 anjea-d S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y £€Q 938ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ (zv'8695 «(¥vLOvT - 03 97'8€S0T -0} €TTOTS N UN 4N EEL -) YE'8TT8- -) vEvSTE- (I %S6) 3 ujeig S|oyM (eT°1T26) (z8°00€T) (6¥°£TOT) M S1I3WINjOA 9'LTTULT- T'T6ETT- 4N | 98'SLLTC- 61'¥6862- 6T°0€05C- (35) @8uey) uean 51 £99 9 4N 291 (44 T N S2W021N0 NN 4N 4N 100">d 434 1000°0>d 4N anjeA-d uN 4N 4N dN (T8) YN 4N | (%)u 'aAnesau piojAwy 4N 4N 4N 434 4N 4N 1D %S6 4N 4N €500 ¥00°0 €0- GZZ'0- | 38uey) ues|\ paasnipy UN 4N 6TT 88 L€ 8 N 4N UN syiuow g jodawig IANS UN 4N 4N 434 1000°0>d UN anjea-d 13d nideaquold 4N 4N 4N UN (S9) uN 4N uopnpaJ a8ejuadlad 4N 4N 4N UN 4N 4N 1D %56 dN dN dN 10°0- 9z°0- YN | 98ueyd uesy paisnipy 4N 4N 4N 96 € 4N N N N syuow T rodawiL (%8°£) 9T $(%¥'C€) 89 uN 4N 4N 4N | (%)u 'AiredaN plojAwy 4N 4N 4N UN 4N 4N 9215 13y UN UN 4N UN 4N uN uoMINpaJ JO U 43 100°0>d 4N UN 4N 4N anjea-d 434 | 09'SS- 03 ¥9°C9- 4N 4N 4N 4N 1D %56 434 65" 4N 4N 4N 4N ues iy CA_%CMMHM ¥9'€ 81'SS- 4N UN 4N 4N (35) @8ueyd uesN $1 66'CC 4N 4N 4N 4N (as) uesy 1502 10TC dN UN dN 4N N syjuow gT 4N UN 4N UN odawil spiojjuad 131d S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y 8€Q 98ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ uN UN UN 7000 8000 4N | @8uey) ueay paisnipy 4N 4N 4N 6€ €€ UN N dN 4N 4N syjuow 8T rodawil 4N 4N UN 43y 10°0>d UN anjeA-d UN UN 4N 4N 4N uN (1D %S6) #a 4N dN 4N 4N dN UN uoI3INpPaJ JuUSIAg 1000 8000 4N 0 S00°0 4N | 98uey) uealy paisnipy 899 819 4N z8 6€ UN N onel 4N 4N 4N syjuow g1 ulodawn] ov/cvdv ewseld 4N UN dN 4N uN UN anjead UN 4N dN UN 4N UN (1D %S6) #a 6'68- 6LTY- UN 4N uN 4N | @8uey) ueay paisnipy TL TL 4N 4N 4N 4N N | |w/3d 'Ob-1dv 45D 4N dN 100> UN 4N uN anjea-d uN 4N UN 4N dN UN (1D %S6) #1a 69 8'18¢ 9'50¢ 4N 4N YN | @8uey)d uealy paisnipy L6 10T dN 4N dN uN N | |w/3d '2p-19V 4D SaWo021nQ ewse|d 1B 1S (enjea 43y 6€00°0=d (606°0) O€€"0 43y (660} ¥Z'0 (000°T) 055°0 _d yauung) snjea-q o : : N N ¢m¢.om-%hwwm.mm W o LL9V-) LWMM 0} S.wmmm.wfi. (12 %g6) 4ia awnjop TS0t 1°981- | (TZEOT) ¥¥999¢C- | L6T'LS'T- (89'¥T) vL'9LC- (S¥'TT) L8'V9T- (35) @8uey) ueay 7 sndwesoddiy 199 €V 91¢ <91 [44 1 N [HIN SL3dWIN|OA (enjea 43 1000°0>d 4N 434 (000°0) T00'0 (z000) 1000 -d \euung) anjea- . o : o awnjop N UN UN 434 ww.wnwwumn%wwmw me*MMw mnm_uwfimfiw (1D %s6) Hia Je[NJLUIA !IHN duBWINjoA L5595 8VEVL UN | €0SvveS | (CTEVE)9¥'299L | (9L°£97) SO'Y0S9 (35) 98ueyd uesy S1 £99 9 4N 19T U T N (anjea 434 1000°0>d 4N 43y (000°0) TO0'0 (990°0) ¥00°0 -d B3auunq) anjea-d S]U3IUO) JO 9|gE] 0} uImay 1oday 93uapIA7 [eul{ - 3seasiq S JaWIBYZ|Y 6€Q 93ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ 4N UN (5°LS) £°85- UN 4N UN (1D %s6) Ha 4N dN %8Y 4N 4N UN uonINPal JUAIAd | wi/3d 'urey) s 00'8L 00°¢S S0v'0T L6V'89 UN 4N | 98uey) uealy paisnfpy | IUBWEJYOININ 15D L6 0T €T 91 4N UN N 434 90°0=d UN UN UN UN anjea-d 4N 4N 4N 4N 4N UN (1D %s6) Wa Jw/3d 'urey) w31 UN 4N uN 4N 4N 4N uoloNpaJ JUBdId JuawWe|yoInaN 6 8T UN UN UN 4N | 98ueyd uesy paisnipy ewseld v.S 6CS 4N UN UN UN N 434 1000°0>d SYT'0 4N 4N UN anjea-d UN UN (5°62) 8°€V- UN 4N UN (1D %s6) Ha /5 uN UN %TT 4N UN UN UOHINPaL BB | 46 190.nN 459 68T 9'TL- 796'9¢- 680°L uN 4N | 88uey) uesy paisnipy L6 ¥0T €2 91 4N UN N 434 100°0>d €700 4N 4N 4N anjea-d UN uN (Ly)eTr- UN uN uN (1D %s6) Wa dN dN %ET UN 4N UN uoIINPaJ JUADI I8TNel-d 4D 6'CT 6'GT- €€T- 4N 4N UN a8uey) ueay 1 86 10T €2 4N dN UN N 43y 100°0>d dN UN 4N UN anjea-d UN UN 4N 4N 4N UN (1D %S6) wa i/8d UN 4N dN UN 4N UN uoIINPaJ JUBIID 'NEl-d ewIseld zo 9°0- UN 4N 4N UN a8uey) ueay 1 609 065 dN 4N uN 4N N 434 1000°0>d L9°0 UN 4N UN anjea-d UN 4N UN 4N UN 4N (1D %S6) ma €66 1°62 88T UN UN UN | o8ueyd ues|y paisnipy 86 10T dN 4N dN 4N N jw/3d ne3-1 4S) UN UN uN 43y 10°0>d UN anjea-d 4N 4N 4N 4N 4N 4N (1D %S6) Wa S]U3IUO) JO 9|gE] 0} uImay 1oday 93uapIA7 [eul{ - 3seasiq S JaWIBYZ|Y 0tQ 98ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ 434 %8E- N UN N N aUl|29p JUDJ3d 434 1000°0>d uN N uN UN anjea 4 (uapanq Joupred - - Apnis) mainuiug 8's 9'€ UN UN UN YN | 3ueyd uesy passnipy uaping ez gsL (494 UN UN UN N N 434 %ET- N UN N N auloap JuadJad . (Axoad 434 S0'0>d N N N UN anjead 3 ; Aq jyuedpipued) €z 8'1- UN UN N N | 98uey) ues|q paishlpy av-100 ¥sz €TL N UN N N N 434 %95- N UN N N aul28p 1U3dJad 434 T070>d N N N N anjea d (wedpied) T g0- UN UN N UN | @3uey) ueay paisnipy av-100 €S2 STL N N N N N 434 %61~ N N N UN BuI[3p BB 434 10°0>d UN 4N UN UN anjead (quedijed) 4 TC UN UN UN N | 88ueyp ues|q paishlpy 15-as-03 ¥sz STL N UN N N N salnsesaiA 9 jJo >u.__m=d paje|oy-yljesH 434 SN £1€°0 UN HN HN anjea-d S1U9U0) JO 3|qe] O} uiniay TvQ @3ed Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y £70C 'M3IA9Y JWIOUODT PUE [EJ{UI]D 0} SINHISUIQ LTL0 6€S°0 ¥29°0 anjea-d «(TOV'T 03 T6T'T-) S6E°0- «({TTT'€ 01 2TV' T-) 80 «{0€6'7 03 985°T-) 2£9°0 (1D %S6) H1a 3103s (96°0) 905t (8z'T)9vL'S (8¢°T) vLS'S (35) s8ueyd uesy 571 | ¥T80D-svav 19 ve €€ N N N N *SA duI[Pap aAIHUS0) u| co_uusz"__ 1060 6550 6S¥°0 anjea-d 24025 gS-4aD +(6€°0 01 015°0-) 9€0°0- «(218°0 03 ¥8€°0-) ¥TZ'0 «(Z€€°0015/8°0°) TLT0- (1> %S6) $a (Sz°0) €91'T (E€0) €TLT (ve0) LTt (3) @8ueyo uesy 1 L9 9€¢ ve N sowodngo w>_"_._=mou syjuow 8T jodawyy 68 8t s N duijaseg Ap1aamiq 8%/3w s 427 Alyuow 38%/3w G €21 Apisamiq 8%/3w 5z 901 swuy Apms 20:120T '€ 19 uosuems 924no0s (ApMS T0Z-000D JO SISOP J2MOT) SWO0INQ S'Ed 3|qel 1P 0€>% '(ogaoe|d) pi€ = u (qewauedd|) ySE =u uone|ndod Apnisqns | 3d pIOjAwy ) 12 %064 10113 paepue)s ;3§ '10119 piepue)s :dsS '9oualael :43Y 'ogaoe|d :0gd 'panodal jou (YN '4aquinu :N 3uidew] soueuosal J11sudew (YW 'wesdofy Jad sweadijw :By/3w 'qewaueds| :g)7 'seaenbs 1ses| 'S '|easalul 32U3pPYUOI 1| 'saxog Jo wns - Suney enuawaq [ed1UlD :GS-HAD '(uolsian DIN) Adoluaau) Sulal] Ajleq Jo SINAIY - Apnis 2Alleladoo) aseas|q S Jawidyzly av-DIN-SIAY 'Wal-pT 912Isqns aAIHUSO) - 9IS JUSWISSISSY ASeasI S aWIdYz|y :HT80)-Syay '94025 ausodwo) Qv :SIWOIAV 'OSeasip S, Jawidyzly :qv (9°sT) * e1J] JO uollenuiuoas) 168/0VT (8'8T) 868/69T UN UN UN UN lem § a UN UN N nmwm (Tv¥) T9T/1L (¥ro€) €92/26 juawiealy Jo uopenuRUOISIA sajey dn-mojjo4 S1U9U0) JO 3|qe] O} uiniay 7va @8ed 1oday a3uapiag |eul{ - aseasiq S, JawWidyz|y €207 'M3IASY J1WOUODT pue [edjul|] 10} SINIISUIQ 1 6€ ve N (z€0°0) OO0 (600°0) TOO'0 (€60°0) S00°0 (snjea-d nauunq) snjea-4 +{06€°9¥CC +(9989/¢C¢€ +(ESO'TETT (19 %E6) id uje.g - 01 886°ZYTL") 689 ¥V LY- - 01 OV8'STT6-) EGE'9619- - 01 GSH'T628-) YSTTTLS- JJOYM (1M (€S ETYT) ¥S'02S9C- (sv90LT) TT'TL6LL- (zz's08T) TT°£869C- (3s) @8ueys ueay S 3113WINjOA sS 8€ € N SoWodIng0 |HIN UN Y 4N anjeA-d UN UN UN 1D %56 102°0- 9€T°0- 660°0- (as) @8uey) uesy ul duaIRYIQ L6T0- TETO- ¥60°0- a8uey) ueay paisnipy ¥ €2 €2 N (4ANS) sewo2NnQ piojAwy 4N 4N 4N 'SA 9U]]29p w>_u._:w00 ul CO_HUJWM" uN UN uN odaoe|d *sA JuawsAosdwi JuadIdd 6580 TL6°0 7650 anjea-d (55070 03 #¥0°0-) 900°0 (T90°0 03 #90°0-) TO0"0- (21070 03 £80°0-) 20Z°0- (12 %06) $1a (€0'0) 66T°0 (v0'0) Z6T'0 (v0'0) €£T°0 (3s) @8uey) uesy 1 19 qE €€ N syjuow 8T jujodawi] SINOOav N N N *SA BUI|IBp |ed1ul]d Ul :o_uusz" T€L°0 ¥15°0 9€€'0 anjeA-d (9%0°0 03 0£0°0-) 800°0 (990°0 03 620°0-) 610°0 (9£0°0 03 020°0-) 820'0 (12 %06) Ha (oz'0) 6€T°0 (€0"0) 6¥T°0 (€0'0) 85T°0 (3s) @8ueyd uesiy S1 L9 w 8¢ N syjuow g1 ulodawi 4N 4N 4N 0qgade|d "sA JudwdAosdwi] JU3IIRY odd *SA dU|Pap dAIHUBOD u| UoIINPAY S1U9U0) JO 3|qe] O} uiniay €y @8ed Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y £70C 'M3IA9Y JWIOUODT PUE [EJ{UI]D 0} SINHISUIQ 81 14" (3s) @8uey) UEBN ST ov 1€ N as-4ad auljaseq 370 01 suljaseq 310D jutodawi| 0 €0 (3s) @8uey) UESN ST 8¢ 1¢ N Syiuow g1 031 aulBseq 310 wjodawi| S0"0>d S00>d anjea-d SINODav 820 8T°0 (35) 33uey ueay S7 oy o€ N auldseq 370 01 duldseq 310) jurodawi] sawo2nQ aAluso) o 8¢ N auijaseg «p3edul-oqaoe|d 310D *MZ0 3/8wQT qewsueds| aseyd a10) suuy Apmis Lor'901'201 TC0C DIUBIDIUOD QYLD '|e 19 uosuems 'ASojoinaN TZOZ '| 19 uosuems {5110202 °[ 19 UOSUBMS {TZ0Z I2UdIDJU0D AV "|e 12 UOSUBMS {,,,220Z "|& 19 apeddIN asinos 310 102-0009 el Sawo02InQ 310 qewaueda] 9°€d 3|qel 13 %064 10413 pJepuess :3S '0L pJepuels :(S '@ouaJajad 43y 'ogade|d :0dd 'paniodal 10U (YN 'Jaquinu N 8ui8ew) asueuosal d11vuSew (YN 'wetso|ry Jod swead|iw 8y /8w 'qewauedd) :g)T 'salenbs 1sel| (S '|eAIaluUl S2UBPIIUOI (D 'Saxog JO wng - Sulley enuswag [e21UlD :4S-HAD 'WaN-T 3]easqns SAINUS0) - 3|BIS JUBWISSASSY 9SESIC S, JawWldyz|y (¥ T80)-syay '94008 ansodwo) gy :SINODAV 'Sseasip s awiayzly :ay (962°0) 6100 (692°0) L10°0 (82€0) 2200 (anjea-d nauunq) anjea-q awn|oA +(VEE'TT- 01 TTV'89) ELT'OV- | +(VEQVT- 03 ¥L6°'6L-) YOE Li- +(6¥S°€T- 01 99€°78-) 856°L1- (1D %S6) #a | sndwesoddiy (SS6°ST) 697" L6C- (€50°6T) 009°0€- (191°02) ¥ST°S0E- (35) ®8ueyd uesy 57 'N SS 6E ve N J119WN|OA (€ETZ°0) €£T00 (T00°0) TOO'0 (069°0) £90°0 (anjea-d nauunq) anjea-4 awnjop +{TLTPS9T +(902°589¢ Je[NILIUIA . o154 . . | 01 TSZVEE) 9L2'V66 0196626TT) T8z-Tzer | *\LLEBOLT O3 SLY'EOT) 926'S06 (12 %56) #1d U (vLT€LE) 6LL7BEE (T8€°svy) s82'992L (v¥8'69Y) V0529 (35) @8uey) uesy $7 JIdWN[OA S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y ¥Q o8ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ 1€ at N syjuow ¢T1 J70 03 auljaseq aJ0) u:_onmrc_._. ST0- SS°0 a8uey) ueajy paisnipy oney ov/zvdv ewseld [43 St N ouljaseq 310 01 auljaseq a10) ujodawi] S2W02INQ 4S) (08) v UN {96)u '@nnnedau plojAwy syauow ¢ N Jurodawi) (€8) ot UN (%)u 'annesau piojAwy syuow gt HN jujodawi] peay |ensjA plojAwy (sL)9 4N (%)u '@Anesau piojAwy syiuow 9 4N u:_On_mE_._. (ev) € 4N (%)u '@nnesau projAwy syuow g 4N julodawi] €T°0- 80°0- a8uey) ueay paisnlpy qT LT N SYJUOIAl ZT 01 duljdseq 310 urodawi) JANS 13d s1deraqiol4 0 9z'0- a8uey) ueay paisnipy 9T 4 N aulPseq 310 01 duldseq 240D julodawi] sawodnQ piojAwy 0t T (3S) @8ueyd uesy 1 14 it N syjuow QT 03 auljsseq 370 u:_onmrc_._. v1800-svav 90'vT ¥5'6 (3s) 98uey) uesy 1 ov 214 N auljaseq JQ 01 auljaseq 240D u.r__onmrr__._. TE v'e (35) 98uey) ueay 1 6C v N syluow gT 01 auljeseq 3710 p:_onmE_._. SJUaIUO) JO 9|geL 03 UIN1ay 1oday 27UapIAT [euUl{ - seasI( S JaWIsYyzZ|y Syq 28ed €207 'M3IASY J1WOUODT pue [edjul|] 10} SINIISUIQ uojpjeso|e Apnis 8403 syl uodn paseq SWLIY, nej B30} :nej-} 'olle. anjeA ayeldn pazipiepuels JANS '4a|lj|iw Jod swei3ooid jjw/3d 'uolleinsp paepuels :qs 'J04Jd piepuels 13 'S23M Jnoj AIBAT :MPD 'Sieom oml A1aaa (D 'nel palejAioydsoyd :nej-d 'AydesSow o] uoissiwg uoJsod :13d 'Uuoisuaixa [sqej-uado 1370 'parodal Jou (YN '1aquinu :N 'weJso|y Jad weadiiw :8y/3w 'saienbs 1ses) :S7 'pinj [euldsouqalsd 148D '|eAIDIUl BIUSPIIUOD 1[I '910 d1sodwo) gV :SINODAV 'plojAwe-elaq gy LT'T- 90°'T- 98uey) uea\ S S¢ oc N syjuow gT 370 01 sujjaseq a4o) julodawi| Jw/8d 'nej-d4 ewsejd 0 7o aSuey) ueay paisnipy LT oc N au||aseq 3710 01 suljaseq 910D julodawi| 8000 L000 asuey) ueal paisnipy SJUaIUO) JO 9|geL O3 UIN1ay yoday 27uUapIA7 [eul{ - seasI( S JawWIsYyzZ|y 9vq a8ed €207 'M3IASY J1WOUODT pue [edjul|] 10} SINIISUIQ 28ueyd 1y 'uoneinap pJEPUE]S 1S 'SHoM In0J AIDAT :MPD 'Soam om] AIdAd :MZD '%S 1sed] 1e Aq ogaose|d ueyy Jo1iaq aq o3 Ayjiqeqoud :(as)) 4d 'OpnuuBew Aue Aq ogase|d o1 Jouadns aq 01 Aljigeqoud :dns 1d '@sop 06d3 241 8uiaq jo Alljigeqoud :((06Q3) 4d '©s0p BAIDBYD |ewixew Sulaq jo Aljigeqoud :(xepy) 4d 'Aljiqeqoud g 'ogased :0€d 'panodal 10u YN 'weaSoy Jod weadjjiw :8x/8w 10949 JUSWI1EEI] WNWIXEW XA 109JJ9 1USW1L3J] XBWP 3Y1 JO %06 15e3] 18 yum uaw|da. asop :06dd '@aualayip :H1d '(ogadeld ueyy 49139Q %SZ) dUJBYIP JuLIYIUBIS Aj[eD1UlD S '|eAIDIU] DIUBSPIUOD (D 'T ulaloudodijody :Jody '94006 a1sodwo) av :SINODAY 8€9°0 9/60 €950 w90 UN UN N | ¥T00| £L00 EvT MZ0 3/3w 01 6L1°0 196°0 98€°0 8T€°0 UN UN N | TT00| #8070 e MPD 84/3w 01 €500 L9YY0 100 100 UN UN N | 9T00| 9ITO L8 MZ0 /3w § L00 9T¥ 0 1€0°0 7200 4N 4N N | 1200 6TT0 8v MPD 3/3w g 8200 91T0 600°0 600°0 UN UN UN | vZo0| ¥ETO 1§ MZD B)/3w 52 - - - - UN UN N | 2100 €IT0 62¢ o4d sjuedpnJed ||y syuow ZT | jujodawiy (@s2)4d | dnsid | (06a3) 4d | (XeIN) Hd ID%S6 | ueaw 1D %56 | as| uesw sanueny) 10431s0d [oa1u0) wody 3ia aulaseg wou} y sisAjeuy Atewd - SINODQY N leloL gewaueda zorSISAleuy Alewlld SINODQYV - sisAjeuy ueisaAeg qewauedd £°gq 3|qel S]U3IUO) JO 9|gE] 0} uImay 1oday 93uapIA7 [eul{ - 3seasiq S JaWIBYZ|Y LvQ 38ed £20 'M3IASY J1WOU0DT puE [ed1Ul]D) O} SINIISUIQ 9€6'0 | 7660 8240 Z6L°0 ST00-01T9T°0- | ¥80°0- ¥ST'001£20°0 | N | 960°0 St MZD 83/3w 01 6/90 | 9560 80°0 8500 9000 01680°0- | T¥0'0- TLT00180T°0 | UN | 6€T0 8T¢ MO /3w 01 0| 68L0 €100 7100 ¥€0°0018L0°0- | £20°0- Z0Z'0019TT'0 | UN | 8ST'0 18 MO /3w g Zsv0 | 8LL0 2L0°0 8700 €400 01 260°0- | 920°0- ¥T2'001860°0 | YN | SST'O LE MYD /3w g SIS0 | 080 90T'0 600 Zv0'0 01S0T'0- | TE0O- €TZ°001580°0 | UN | 6vT°0 8¢ MZD /3w 5°Z - - - - 43y 43y 917001 ¥¥T0 | N | 810 891 osad sialue) 3 m0fl< 90| 160 Z5°0 £09°0 100°0- 01 €60°0- | £¥0°0- 09T°001060°0 | N | 9zT°0 ZsT MZD /3w 01 €150 | £26%0 Y6T°0 SST'0 110°0 03 2L0°0- | TE00- TLZT'0OIETTO | UN | 2¥T0 vt MO 3/3w 01 €810 | 2790 6100 8100 €0°0 01950°0- | £00°0- S0Z'001£ZT°0 | UN | S9T°0 68 M0 /3w g Z€0 | 61L0 6210 £60°0 I¥0'0 01 €£0°0- | 9100 907°00180T°0 | ¥N | 95T°0 8y MPO 8/8w g €EE0 | 0LO 8€T'0 8710 Sp0'0016£0°0- | LT0°0- 0TZ'001TOT'O | ¥N | 9ST'0 zs MZD0 33/3w 5°Z - - -~ - 434 43y Z0T001ZYT0 | UN | 210 8€T osd sjuedpiued ||y syjuow g1 jujodawi (@sd)4d | dnsad | (06a3)id | (xew)ud 1D %S6 | uesy 1D%S6 | as | ueaw mfl_u._u.:m-,_d 101493S0d JoU0) w0l Hid Jujjeseg wod} y sisAjeuy Alepuodas - SNOJAY N |e301 gewaueda zorSIsAleuy Alepuodas SINODAYV - sisAjeuy ueisaAeg qewauedd 8'gd a|qel S]U3IUO) JO 9|gE] 0} uImay 1oday 93uapIA7 [eul{ - 3seasiq S JaWIBYZ|Y 8vQ a8ed €£70C 'M3IA3Y J1WOU0DT pUe [eJJUl|D JOJ BINUISUID UN UN UN UN UN UN dN | UN UN uN MO B)j/Bw g dN 4N UN 4N 4N uN 4N | 4N dN 4N M0 3%4/3w 57 4N 4N 4N UN 4N uN dN | 8N uN dN o1d siaue) 3 Jody UN | 7960 0 zs0 860°0010/8°0- | €IV O- VLT TO}T8Y'0 | UN | SE80 ST MZD 3/3w 01 N | T6°0 ¥0€'0 s¥T'0 €80°0 0} E€L°0- | STEO- 907'T 03 /£9'0 | N | €260 9T MO 8%/3w 01 iN S9°0 LZ0°0 8200 ¥6£°0 03 095°0- | 60°0- LVSTOIE6L0 | UN | LSTT 68 MZD /3w g UN | 6050 9%0°0 1€0°0 209°0012SS°0- | 2000 LLLTOYERL'O | UN | STT 8 MPD /3w g 4N vL°0 €810 9LT°0 ¥0%'0 03 208'0- | S6T°0- 6/5°T 010250 | UN | €S0'T 4] M0 3%/8w 5T iN - - - BE 434 EVSTOIZS6'0 | UN | 8YTT 8€T ogd sjuedpjued ||y syjuow QT (as3)4d | dnsid | (06a3)4d | (xew)ud 1D %S6 | ueaw 1D%S6 | as | ueaw mo_u_u.-._m:d 101193S0d jodjuo) wiod} Jid gujjoseg woll y as5-34ad N |e101 gewauedd z019S-40D - sisAjeuy ueisaAeg qewaueda 6°€d a|qel 98ueyd 1y 'uoneirsp pJepuels :@s 's)esm Jnoj AIBAT :MPD 'SHoam oml AIDAS :MZD '%S 1ses| 1e Aq ogade|d ueyl Jo11aq aq 03 Alljiqeqoud :((as)) 4d 'SpnuuBew Aue Aq ogade|d 01 Jouadns aq o3 Ayljigeqoud :dns 4d '9sop 0643 Y1 8uiaq jo Ayljigeqold :(06Q3) 4d 'OSOp dA1DDYD [ewixew Suidq jo Aujiqeqoud :(xep) 1d 'Aljigeqoud :ud 'ogadeld :04d 'panodal jou :yN 'wesdofy Jod wed|jiw :8x/8w 10949 JUBWILI} WNWIXEW XA 10513 JUDWILII] XBWp 3Y] JO %06 1Ses] 18 Yum uawidal asop 0643 'Dduasaip Hid '(ogade|d ueyl J9119q %SZ) dudIDYIP JUBIIUBIS Ajlea1ul|d :QSD '|BAISIU| DdUBpYU0) 1D '3 uidloudodijody :3ody '910§ aysodwo) gV SINODJAY 620 €9°0 vLT0 €0 950'0016L0°0- | TT00- ¥£1°001S960°0 | UN | SETO L0T MZ0 3%/3w 01 LST0 TES0 97T 0 81¢°0 6£0°001580°0- | €00°0- S0Z°001780°0 | UN | €¥VT°0 8¢ MpD 3/3w 01 9910 98€°0 9.0°0 7800 €IT003¥L00- | ¥T00 SYZ°001¢260°0 | UN | T9TO 8 MZ0 3)fBw g vz 0 870 ST1C°0 T6T°0 960'0 01 060°0- | 2000 9¢T'001+L00 | UN | 6VT0 1T MpO 3jf3w g €C0 o 6070 6070 TIT'001¥60°0- | 8000 €¥C°001890°0 | UN | PST°O VT MZ0 /3w 52 - - - - 434 434 TOZ°'001¢260°0 | UN | 9¥T0 oL odd sidiiied-uou 3 Jody S]U3IUO) JO 9|gE] 0} uImay 1oday 93uapIA7 [eul{ - 3seasiq S JaWIBYZ|Y 6¥Q 25ed €£70C 'M3IA3Y J1WOU0DT pUe [eDJUl|D JOJ SINUISUID 4N 4N UN uN 4N UN 4N | HN UN 4N MZ0 /3w 5'2 UN UN UN iN UN UN dN | 4N UN UN odad siolle) 3 mon< UN | 8860 LT80 1980 €£7°0-01S6L°E- | TTO'T- TS6'TOIYEZ0 | UN | TI9T TSt MZD 3/3w 01 4N | £18°0 8900 6%0°0 G58°001967°C- | TTL'O- 910'7 03808'T | N | 16°C 9z MPD 33/3w o1 N | 8990 90'0 LEO0 €8Y'T101662°C- | TTVO- 0SLPOITTLT | UN | TTTE 68 MTD /8w g UN | 9Z¥0 1€0°0 7200 €EET010E0°C- | 8ZTO TZ6'SO10T6'T | UN | 6S8°€ LY MPD /3w g 4N | 8Zv0 8€0°0 €0'0 €EE'TOMYIT'C- | STTO L00'90I8YL'T | YN | LSS'E s MTD 3 /3w 5T UN - - - 434 434 99/'703T0SZ | UN | ZE9°€E LET odd sjuedpnJed ||y syjuow gt (@sD)4d | dnsid | (06a3)id | (xew)ud D %56 | uea 1D%s6 | as | ueaw sa1uenY J0113150d |o41u0) Wo4y JHd sujoseg wody y ¥180)-svav N lezoL qewauesa 20:7T802-SvayV - sisAjeuy ueisaAeg qewaueda] 0T €d d|qel 98ueyd 1y 'UoIIBIAIP pJBPUE]S (]S 'S)23M JNo} AJBAT (MO 'SHPaM oM AJBAd :MTD '%S 1sed] 1k Ag ogade|d ueyl Jansq aq 01 Aljiqeqoad :(as)) 4d '@pnuuBew Aue Aq ogaseld o031 Jouadns aq o1 Ayjigeqoud :dNS 4d '@sop 06a3 2yYa 8uiaq jo Aujiqeqoud ((ped3) 4d 'Osop DA118Y)9 [ewixew Sujaq jo Alljiqeqotd ((xeAl) 4d 'Aljiqeqoad ud 'ogade|d :0dd 'parodal Jou (YN 'wea3o|n Jod weadijjiw 8 /8w 10949 1UDWILDI] WNWIXEW :XBA 10349 JUSWIEdL) XBWP dY] JO %06 ISEI| 1B YUM UswI8a. 9sop 063 'Douataplp :4iq '(oqaoeld UBY} 131194 %5Z) 22U JUedUS|S AjEDIUlD :dSD '|BAISIUI SOUSPUOD ([ 'sBXog Jo wns - Sujiey ejipuawaq [edul) :gs-4ad '3 uieroadodijody :30dy UN 4N UN 4N 4N UN 4N | ¥N UN UN MZ0 3/3w 01 UN 4N 4N |\ 4N 4N 4N | UN uN UN MO 33/3w 01 UN 4N 4N 4N 4N uN 4N | 4N UN UN M0 /3w g UN 4N 4N 4N 4N UN 4N | 4N UN UN MPD /3w g dN 4N UN uN 4N uN 4N | UN dN UN MZ0 3/3w 5'2 uN 4N UN 4N 4N UN UN | 4N UN uN odd sJallied-uou 3 Jody dN 4N UN 4N 4N dN 4N | UN dN dN MZO 3/3w 01 UN 4N 4N 4N 4N 4N 4N | 4N UN UN MpD 3/3w 01 uN 4N UN UN 4N uN 4N | HN uN UN MZ0 /3w g S]U3IUO) JO 9|gE] 0} uImay 1oday 93uapIA7 [eul{ - 3seasiq S JaWIBYZ|Y 05Q 98ed £20¢ 'M3IA3Y d|WOU0I] pue [edul|) J0} SINYISUID uN 1€ (0)o (0)o (0)o (0)o (0o (0)o Yieap 03 Sujpea I-viyv (80)L {£0)9 (80)¢ (0o (80)¢ (Tort (0o (8€)e Yyieaq . . (s92) . . . . . (6'18) S€L (6'88) 86 oo | EivIoL| (resiser | (Lee)ie| (0'6v)sT (zvv) €T 2ViL patejaiuawIeal) (6'18) S€L (6°88) 86L (e Mw (e98)6€T | (Tv6)8ET | (0'88) T8 | (T'v6) 8P (s88) 9v 3udA3 3s1anpe Auy SJUIAJ 3SIOAPY L68 «368 S 19T €9¢C [43] 1§ [4] N Apjaamiq Apjeamiq Alyuow ApjPamiq Alyuow Apjeamiq ogaoe ogaoe suly Apn A998l SyBwor | *7°M | syBwor | SyBwor| ByBws| SyBws| SyBwer v Apms h.__.__.U Cfl___& 'g11€20Z IUBIJU0I aV1D 'e 19 9¢€20T BIUIBJUOI AVLD 'y,;220T AdpueT £, T2ZOT "|& 39 uosuems yseqqes ',.zZ0z "Ie 19 YoAQ uea 92In0S av ALuvD 102-0009 el Aiajes TT°€qQ 3|qel a8ueyd iy 'uolielAap pJepuels :dS 'SYaam Jnoj AIRAT I MPD 'SHoaM om) Alaae IMTD '%S 1sed] e Aq oqaoe|d uey) Janaq oq 03 Ayjiqeqoad :(gs)) 4d 'OpnuuBew Aue Aq ogade|d o3 Jjouadns aq o} Ayjigeqouad :dns 4d '"@sop 06a3 3Y3 Suiaq Jo Aljiqeqoud :(06Q3) 4d '950p AI0344D |ewixew Bulaq jo Aljiqeqoad :(xe) ad 'Aljiqeqoud tud 'ogadeld :0gd 'parodal 1ou tyN 'weaSo|iy Jad wead||iw 8y /3w 109)J3 JUBWIEDI} WNWIXEW XBIA 10549 JUDWIEDI] XBWP BY] JO %06 ISed]| 18 YUM uawidal 3sop :06A37 '@dualayip JHa '(ogadeld ueyl Jo119q %S7) 22Ualaylp jueayiudis Ajleatul)d :as) 'leAdsiul duspluod i|d '3 uisloidodijody :30dy 'WRN-HT B|BISANS dAINUSOD) - B|EIS JUBWISSASSY Bseasiq s Jawisyzly :#T180d-svay 4N 4N UN dN 4N 4N 4N | 4N dN 4N MTZO 3/3w 01 UN UN UN UN UN 4N 4N | ¥N UN dN MPD 3/3w 01 UN 4N UN uN 4N UN 4N | UN UN UN MZ0 3/3w g dN UN UN UN UN 4N 4N | 4N UN dN MPO /3w g UN UN UN UN UN 4N 4N | 4N dN dN MZ0 /3w 52 UN 4N UN dN 4N 4N 4N | UN 4N UN osd siollied-uou 3 mon< UN UN UN UN UN 4N 4N | UN UN UN MZD 3%4/3w 01 4N 4N UN dN 4N 4N 4N | UN 4N 4N MrO 33/3w 01 UN UN UN UN UN 4N 4N | 4N UN dN MZ0 /3w § 4N 4N UN uN 4N UN 4N | HN UN 4N MPD 34/3w g S]U3IUO) JO 9|gE] 0} uImay 1oday 93uapIA7 [eul{ - 3seasiq S JaWIBYZ|Y 15Q 28ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ suoieal pajejal (e'TT) TOT (pzr) 11T 4N 4N UN UN UN 4N ~UOISNU] 10 YTHY INOYYM ST SNOSS (tTot (970) s UN UN UN UN UN UN H-VIaY 0 (80) L uN UN 4N 4N uN UN VY 0 (zT) 1T UN uN 4N uN uN uN suondeal pajejaJ-uoisnyul SJV snoLes 4N 4N VN %0¥'ST 4N 4N 4N uN aA1sod VN dN UN VN %060V 4N uN uN UN (vav) seipoqnue Snup-nuy (£9) 09 (T2)vo 4N UN 4N UN 4N 4N 61-AINOD (8's) TS (£°9) 09 4N UN 4N UN 4N 4N ujed yoeq (Tv) 8e (0's) sp UN UN UN 4N UN UN Aanxuy (59) 85 (€5) 8v uN uN 4N UN uN 4N eayulelq (T'6) 28 (£8)8L UN 4N 4N UN UN UN in (679) 79 (69) €S dN UN 4N 4N uN UN ejesyuy (s2)ze (9s) 0s 4N uN 4N UN 4N UN SND jo sisoiapis [epysadng (T8) €L (T'TT) 00T UN UN 4N uN UN UN ayoepeay (T's)op (s's) 6¥ UN UN UN UN UN UN ssauzziq (9°6) 98 (vot) €6 uN UN 4N UN uN 4N led %SL UN UN UN 4N UN UN (asop 1s114) uopdeal parejas-uoisnju| %96 UN UN UN 4N 4N dN | (2 Jo T apesn) uondeas pajejas-uoisnjuj (¥'2) 99 _ §(v'97) LET | %OE'E %06'6T %06°C %00'CT %08°L %08'S uondeal pajejas-uoisnyuj SJUDAZ BSIDAPY JDYIO UN UN UN Swmm mm%mmh (ee)lee/e | (0T)1s/T | (6T)Es/T 5/ 03 Buipea) 3-viuy (T€) 8¢ (£s)ts | (T9)sT (evT)ve | (98T)iv | (60T)0T (80¥ $SeT) L lel jo D/q o1 Buipes| 3y juswieal jo uN N | %080 %0S°C 4N dN uN N | /a 03 Buipes] uonoes: pajejB-oISNIul (627) 9¢ (69) 29 4N 4N 4N 4N 4N uN juswieaJy Jo J/q o3 Bujpes] Iy (€'11) TOT (ovT) 92T G.nmfi_w (estise | (ger)1e| (viT)ot (8L ¥ (zet)oT JUDAD 3SIBAPE SNOJIAS uN (0)o (0)o (0)o (0)o (0)o (0)o yreap o1 Suiped] H-vVIuY S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y 75Q 9@8ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ . 9108Az013)2y aanisod 8/v/0 (1) 621/8 UN UN UN N UN N ¥30dY neWoIdwAS aAnsod I-vIyv > T19/0 (v'e) 0z9/1T N UN HN N N UN 7304V :onewordwAs newoldwAg 0 #(87) st N (zT) e (ro)t (TT)T (oo (6T)T anewoldwAg % JEWIURID| yum syuow 9 Aq %26 'shep 0ZT VRV N Aq %T8 'shep 06 UN UN HN UN UN N 40 uonn|osal a1a|dwod Aq %9 'syjuow 0} swi} ueipa|y ¥ ulyum uonn|osal %I8 VN € HN UN UN UN UN HN DIDADS :AJIDABS [BIIUID oleJapowl VN (4" UN N dN UN UN N -A11I3AS (DI VN ot UN UN UN N N UN Pliw :AJLIBABS [E2IUID 919A0S 0 6 UN UN UN UN UN N -Kianas oydesSoipey 9jeJapow I-Viuvy 9 99 UN N dN UN UN N -Kauianas aydeSorpey 6 LE UN UN UN UN UN UN Pl :Ajuanas a1ydesSoipey (tTo)t (Te) 8t N N dN UN N N 3-VIHV Jua.1nday (€'0) 982/1T (r's)8sz/sT N%M (vzeat/e| (oztt/o| (o) zit/o| (o) zTiT/O {o)ztT/0 aAnedau $30dv . 9108Azowoy (8/¢€) €t/s (9'z€) TPT/9v UN N UN N N N samsod y30dy . . 9108Az01919Y (6'T) 8L¥/6 (6'0T) 64%/25 UN UN UN 4N UN UN samsod y30dy (e2) TT9/PT (8'sT) 029/86 aw\w (v¥)ot/L | (z6)sz/ez | (ooT)€/€ | (0OT) T/T (ooT) T/1T anisod y30dvY (8°0) (6°6) (6'6) . . . . » >C (£T) 8T (9zT) €TT evz/z 191/0T ccz/st (ee)ee/e | (0 TS/T | (6°T)2S/T v SJUaA] VYUY S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y ¥#30dV :ohewoldwAs €6Q 9@8ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ (zo)z (zo9| (0o (0o | (0o | o] (o] (0)o dneworduwiAs (zv) 98z/TT (6'TT) 8L2/5€ UN «+(9'P) ELT/8 aAnedau $30dv . . 9108Azowoy (T'T2) €€T) 8¢ (0'6€) T¥1/SS UN UN UN N UN UN samsod y30dy . . 9108Az0J913Y (9'8) 8£v/T¥ (ovT) 6L%/L9 UN N UN N UN N samsod y30dy (€TT)T19/69 | (L'6T) 029/zzT N «+(T'€T) 9EV/LS anmsod p304y | (MY HVIIV (o)t (970)s @%W {(9°0) T9T/T HN N N N agey.liowayoioen (ea)te (975) 05 UN N UN UN UN N s1S0JapIs |eloyJadng (9°2) 89 (ov1) 92T N UN HN UN UN N 98eyJiowayodiN (0'6) T8 n(g/T) ST | (€'S) €T (89)TT (TTr)87 | (S8T)/T (L'€T) L (8'€)e Auy (1) aAe3au 98¢/0 v'T)8L2/Y N N UN N N UN y30dY onewoldwAs 9108Azowoy aanisod £€T/0 (z6) TrT/eT N UN HN N N N } 4 Snal S1U9U0) JO 3|qe] O} uiniay 1oday 2ouapIA |eulq - aseasi(q S JowIayz|y 5@ a8eq £70C 'M3IA9Y JWIOUODT PUE [EJ{UI]D 0} SINHISUIQ *S9SOP qRLWBUBID [|B SSOJIY 44 *AjJed pa.Ind20 3-y[HY YUm H-VIHY Ing '9S1n0d Juswieas) Sunp Ajwopuel palinado H-yIdy jo Suiwijx *uoISNJUO3 pue 'aduequnisip |BNsIA 'Sydepeay a19m swoldwAsy *suopoeal palejal -uoisnjul Joj (Sp102110202n)|8 Jo 'saujwelsiyiiue 's8nip Alojewiwejul-ijue [eploJ3lSuou 1) SUOIIRIIPSW dAIIRIUSASId 3)e) Jou pip syuedidided ay) Jo %958 'HodaJ ujew ayy wodj sanjeA wodal apn "1uawa|ddns pue 1dpasnuew Ul JUIISISUODU] SeM sah|eA asayl Jo Suiioday: v 258Ud 370 pue 8100 Y3 ul pariodayy uonieindod Ajajes, uoJ193)u] Joe41 Aleupn LN 'uoilddjul 1oea) Aloleaidsal Jaddn | yN 'SIUSAD dsiaApE JuSBIswWI-udWIeaN] 13yT] 'ogade|d :0dd 'panodal jou :yN 'Apoqiiue Suizijesinau :qyN '9|qedidde 10U yN 'wesSon 4od weadijjiw :8%/8w 'uollenuiuodsip :D/q 'WalsAs SNOAIBU |BIIUD (SN 'SISOISPIS |e1d1aa3dNns J0 a8eyioway 03 anp sallljewloude 3uidew) pale|aJ-plojAwe (H-y¥[YY 'uoisnysa/ewaps 01 anp ssilljewsouqge Suiew| palefaJ-plojAwe :3-y|yy '3 uisloadodijody :Jody 'gewnuednpe QY 'WSAS SSI9ApE ]y (8°2) 0L (68) 08 4N 4N 4N UN 4N UN 3-VI¥V INOYUM H-V 4V (0T)6 (Teve | (o)t (V)L (ra) et (T1)T (0ot (0o H-VIV Yum 3-Vidv (s'6)s8 (s'T2) €61 UN UN UN UN UN UN H-VIYV 10 3-VI4V (T0) L68/T 0 UN 4N 4N 4N UN 4N | 98eysiowsyoioew [e1gaiad Yum syieaq (zo)z (0)9 UN Y UN Y uN uN aSeyuiowayosoew [RIGDII) S1U9U0) JO 3|qe] O} uiniay GGQ @3ed 1oday a3uapiag |eul{ - aseasiq S, JawWidyz|y €707 'M3IASY J1WOUODT pue [edjul|] 10} SINISUIQ 'snoBAzoia18y Y% T'TT pue snodAzowoy %S¢ 'syuedidiyed pajeasy-ogase|d 40d pue $3 Jody,, JUDAD 9SJOAPE JUDSIBWS-JUBWIeAL) :3y31 'snoduelnaqns :)s 'Apoqiiue Sujzijesinau :qyN 'parodal jou tyN 'Sjqedijdde jou :yN 'Jaquunu :N 'SNOUIARIIUI (A] 'SISOISPIS [eIadNs Jo 38eysioway 01 anp sanljewJloude 3uidew! pale|al-plojAwe (H-y|yy 'Uoisn)ja/ewapa 01 anp sallljewsouqe SuiSew pajefaJ-plojAwe :3-y|yy '3 uisloidodijody :Jody VN {(9'0) 08T/T asSeyliowayoe H-YIuV (%0) ¥T/0 uN aAleSau 3 Jody «{%T'CT) €€/ 4N aAllsod §3 Jody I-Viuvy %06'8 %08'L Auy VN %000 9ARIsod qVN VN %019 (vav) seipoque Snip-nuy %09°0¢ uojjoeal pajejal-uoisnjuj pajeasy-ogaseld s10) Apjeamiq B /3w o1 suy Apmis str60t€C0T '1€ 19 uewapAay 'zzoz Apue 3%inos 310 102-0009 lenL %L0°0 %€0°0 (vav) ssipoqpue Snip-puy %0€'EE (z 10 T Spean) uojoeal paje|a-uoisnju| %0.°0C VN uondeas ays-uondalul %L0°0 %€00 3v3il palejal-juswiieal] 6¢ o€ N uopaaful IS uoisnjul A| swuy Apms 211€C0C "|e 19 |emey 924nog JS -T 9seyd :qewduelo leny Kyayes [euolippy ZT°€d 3lqelL S]U3IUO) JO 9|gE] 0} uImay 1oday 93uapIA7 [eul{ - 3seasiq S JaWIBYZ|Y 95Q a8ed £20¢ 'M3IA3Y d1WOU0D] pue [ed1ul]) JO) SINNISUIQ 89/°0 (z18'00199T'1-) £LLT°0- (00s°0) 8¥6°'T 68/8¢ av piln MZO 8y/8w § 69.°0 (80%°0 01 585°0-) 680°0- (99Z°0) 80°T 68/6€ av o1ang P : '70 I- : : : 1 SIsAjeu 1650 (z20'T 01 ¥50°T-) 605°0 (£88°0) ¥e9°C 8/9 av piin MO 83/8W § syluow gt zor(sishjeuy 9610 (86£°001TEE0-) €£7°0 (eTe0)2ov'T 8/0€ av o1ang P WYWIN) TOZ-0009 688°0 (8sv'T 01 ¥2Z'T-) LTT°0 (ovL'0) TvT'e T5/1T av P MZ0 /8w §Z 7970 (z61°0 01 ¥10°'T-) TIV'O- (Tve0) L5270 7s/¢€t av o1ang D gewsueda anjepd | (12 %06) 110 NS _ (3s) ns1 | N sdnoJ8qnsg swiy Apmis Jurodawil len1 gs-4ad Aaaas aseasig Aq g5-4a) 11 dseyd qewauedad - eyeq dnoadqns ¢1°€q a|qel J0.1J3 pIBPUEBIS 35 '90UDJ3JR 4TY 'SH9M INOY AJAS IMPD 'SH9aM oMl AIBAs (M ZD '0qade|d :0gd 'Jaquinu |e301 /1aquinu :N/u 'welsoy Jad weJs)[jiw :8y/8w 'quawiedw] SAIHUS0D pliw (DA 'UBSW s3J1enbs 1se3| IS '@2UaJaIp JId '91005 axsodwo) gV SINOIAY '9seasip s Jawisyzly :av 434 43y (9g00) €20 8ez/6v av P 08d 434 43y (8T0°0)69T°0 | 8EZ/TTIT | AV 0IanA D ¥1°0 (6000 03 TLT°0-) T80'0- (€¥0°0) 6¥T°0 tst/ee av pin MZD 8%/8w 0T 8500 (£00°0- 01 SOT'0-) 950°0- (szo0) €TT0 2ST/Lv | av oaeng P S19°0 (850°0 01 60T°0-) 920°0- (6€0°0) S0Z°0 ave/sy av pil MPD 8%/8w 0T LTT0 (0100 03 890°0-) 620°0 (610°0) ¥1°0 | 9¥Z/TOT | AV 033nA DN - 9.0 (6TT°0 01 ¥80°0-) 8T0O'0 (zso0) 8vz'0 68/v¢ av P MZO BY/Bu § 4 80 (9%0°0 01 £90°0-) 800°0 (620°0) 9T°0 68/L€ | v 013na DN €610 (622°0 03 560°0-) £90°0 (z60°0) L6Z'0 8v/S av Pl - - - - - - MPD 8%/3w § L0 (£¥0°0 039£0°0-) ¥T0°0- (v€0°0) ¥ST'0 8¥/0€ | Av o1and PN L6E0 S0Z°0 03 990°0-) 690°0 SL0°0) €0 Ts/11 av P A i A i / Pl MZO 3y/8w 5z zor(SIsAeuy ro (¥00°0 01 8Z1°0-) 290°0- (£€0°0) 90T°0 7s/z | av orang PN IWHIAIN) T0Z-0009 gewaue'd| anjep-d | (12 %06) $1d WS (35) W1 N/u sdnoJ3qns sway Apnis ujodawil ledL SIWoDav Aanas aseasig Aq SINODQV 11 @seyd gewauedd] - ejeq dnoigns £1°€d diqel S]U3IUO) JO 9|gE] 0} uImay £5Q @3ed Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y £70C 'M3IA9Y JWIOUODT PUE [EJ{UI]D 0} SINHISUIQ 10113 pJepue)s :3S '9oualaal :43Y 'S)99M Unoy AIAD (MPD 'Sieam oml A1aAs :MeD 'ogadeld :0gd '1aquinu |e1o] /1aquinu :N/u 'wesdopy Jad wedijiw 8% /3w 'quawiedw) aAu80d PlIW DA 'UBBW SaJenbs 1Se3| :AIST '@IUBILIP JHQ 'DSEISIP S 3WIBYZ|Y :QV 'WaL-YT 3]easgqns aAIHUS0) - 9|BIS JUBWISSISSY aseas|( S JBawidyzly :y180)-Svay 43y 434 (9TZ'T) 98°S 8€z/9v av pi'w 084 43y 434 (¥89°0) €9V 8€¢/¢11 | avoiang PN LTT0 £6£°0 01 0/T'G-) 88T°¢- 9Zy'T) ¢LT'E 7s1/2¢ av P - A - - ) - A - ) - / P MZD /8w 0T 97100 | (558°0-01655+-) L0L'C- (896°0) ST6°'T TST/LY | av orand DN 5750 G78'€ 01669'T-) £SO'T v6T'T) vZv9 9vz/sy av pin : A - - J - ( .v ; / P MPO 31/8w o1 997°0 (zsv'00198¥'2-) ZOO'T (zzs0) €9°€ 9%Z/T0T | Av o1and DN Syuow g L18°0 (988'2 01578°€-) 69Y°0- (PEL'T) 168V 68/¥¢ av pi'n MZ0 B8 S 4 8€9°0 (LL¥'T 01659°C) T65°0- (611T°T) 2P0V 68/L€ | av o1and DN $/9°0 6699 01 086°€-) 09€'T VS0'E) 729 8v/S av piin - A - - i - A - ) - / P MPO 8)/8w g 61L°0 (0T6'z 01 0¥8'T-) GES'0 (tec'T) 89T'S 8¢/67 | av o1ana DN IET°0 (zes'8 031 £9€°0-) €80V (69¥°2) evv6 Ts/TT av pi'n : : : : : : / MO 84/8w §°C zr{SisAleuy 98%°0 (0S¥'T 01 T/5°€-) T90°T- (8Z¥'T) TLS'E 7S/Te | av oirand DN INYWW) T0Z-0009 qewoueda aneA-d | (1D %06) 1A WS _ (3s) Ws1 _ N/u sdnoJsqns swuy Apms Juodawi) euL v180)-svav A1aAas aseasig Aq ¢yT80)-SvYQAV 11 @5eyd gewauedad - ejeq dnoi8qgns ST €Q 9jqel 1013 PJEepUB]S 13§ 'D0UDJD)DJ 14TY 'SHOOM IN0J AIDAD I MPD 'Sy9M 0Mm] AIdAS :MZD 'ogade|d :0gd 'Jaquinu [e1o] /1aquinu :N/u 'wesdo|n) Jod weisi|jiw 3 /8w quawiedw 3AIIUS0D plIwW (DA 'UesW saJenbs 1Se3) (ST '93UIBHIP A 'SOX0g JO WNS - Sulley eluswaQ [ed1Ul) :gS-¥aD '9seasip s swisyzly :qy 4134 434 (rse'0) S2T°T 8€z/6v av piin o8d 413y 43y (Z9T°0) 89T'T 8€Z/CT1 av o1ang PW Y900 (86T1°0- 03 £96'T-) €80'T- (91t°0) THO'T ¢ST/S€ av pi'n MZ0 B/Bw 01 LSS0 (987°0 01 £09'0-) 6ST°0- (ogz'0) T0°T IST/6% av o1ang P S/T0 (082'001 €£££'T-) L¥S'O- (£L€70) 85T 9vz/8t av pi'w / MPD B1/8w 01 859°0 (z9Z'0 01 5SH'0-) 960°0- (t£1°0) TL0T 9%Z/101 av o1ang P S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y 85Q 28ed €707 'MIIA3Y d1WoU0d3 pue [ed1ul]D 10} 9INHISU|Q Jaqunu :N 'aseasiq s, Jawisyz|y uo sjeu3 [eauld :av.ld 10 SE0'0- Evy 'CT Japew Ae.s [e911100 3jOYM 160 €00°0- GET 'TET [end|320 SC0 670°0- GET 'TTT |elslied €10 vE0°0- SET 'C€T aje|n3up) en{220T 2uaI33u0d 220 €20°0- GET '22T |ejuou4 avid) av ALV 9700 S90°0- 9€ET 'TET [eJodwa) <100 TLOO- SET 'TCT JeJodwal elan 2000 890°0- SET 'TTT |esodwia) [elpS N anjead ASojoysed ney gewduedd N 'ogade|d N uoiday uieug leML uj 3uUBI3YIQ U3 paisnipy uoi3ay uleig Aq ASojoyied nel Qv ALIRIV1D - ele@ dnou3qns 9T°€q 3|qel S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y 650 93ed €70 'M31A9Y dlWOou0d3] pue |edjul[) JO} 91NUISUIQ 9I¥ 'TI¥ £V €0 S[eiN xas Evv 'vov T o Slewad 9€T 'TET - 870 9108Az0woH 95y 89Y oe s0 9308A201533H snieys adAjouas ¢3 Jody 192 SLT 147 SL0- JalJed-uoN Y0z '€12 i €€0- adoun3 TrT '9v1 S¢ EE0- eisy uoigay ¥1S 'o1S V€ ¢S°0- eJlJawy YuoN L9T°SLe 147 SL°0- J9l1Jed UON sniels 3 Jody 76S '009 1¢ €€°0- Jauie) T€E TEE LT ¢9°0- av PN dnou3qgns |eawuld 8¢S 'v¥S 8¢ SE0- DN Tty L0y 8z 6E°0- oN (ezoz "je suljeseq 19 )2AQ uea) LYY '89Y ST 8v°0- S9A ]e uoljedipswu gy s1rewoldwaAs joasn av ALY gqewauedd Stjuotl gt . Suimols % 1e gS-y¥d)D ul dudBKIA sdnoJ8qng lenL N 'oqaJe|d N uea\ paisnipy oeSdN013qng snouep Aq syjuow 8T e gS-4AD AV ALINVID - eleq dnos8qns 8T°€q 3jqel Jaquinu N 'quadiad :9 'saxog Jo wng - Suney epusaWAQ [Bd1UND :FS-HAD 'OSeas|q S JaWwIsyz|y uo s|euy [esul) :avld 9€ VL0 L0'C 08 'tL sjue|ngeodjuy 8€ £5°0- 8E'T SPT '9€T Ayisaqo [43 S9°0- SLT £€G ''8TS ejwajosaisajoydiadAy (zzoz 9duaiapuod ZS 88°0- L'T T 'TET 9seasip JeaH aviyei €€ 85°0- SLT 0T '0€T sajaqeiq | US4OJ) AV ALIMVID 0€ Ly'O- 95T 1LY '98Y uolsuapiadAy Lz St'o- 991 658 '.8 llesaA0 Suimo 0(32e|d "SA 3IUIIIYI ulPaa ewaueda N 'oqade 3sn Jueindeoonuy e IMmoIs % qade|d Ha 0ga2e|d 9S-4aD q TN '0q3Je|d N pue sappIgJowo) lell asn jue|nSeodijuy pue sanpiqiowo) Aq 9S-4aD AV ALIIVI) -eieq dnou3qns £T°€d 3|qel S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y 09Q 28ed €707 'MIIA3Y d1WoU0d3 pue [ed1ul]D 10} 9INHISU|Q €St '19Y €2 8Z'T- | 9108Azou913H 997 '€LT Q€ 6T'C- J9111ed-uoN shiels adAlouas {3 Jody 20T 'Tt ec TLT- a2doing ovT 'ovT S 8€'T- eisy ¢IS V1§ 1€ £7'T- | eauswy YuopN uoiSay 997 '€LT S€ 6Tz J3114ed UON (zzoz |2 13 885 '665 1¢ IT'T- Ja1e) snieys 3 Jody | APAQ ueA) av ALI¥VID 6Z€ 'O€EE 0€ ov'C- av plln S¢S 'TS €2 €6'0- DA dnoJ3qns [eau) 607 '90v 91 ¥9°0- ON aujaseq 1e Sty 99y 6¢ 90T S3A uonedpaw qy dewoldwAs Jo asn ewauedd Suimo syjuow e y180 h.onmum_n_ .__,_ .* ® -syav :_Muu:m._wu.fl :"HS_ Wmum:.__u< sdnoi3qns m oeSdnoudqng snoliep Aq syluow g7 1e $130)-SYAV AV ALIYVID - eleq dnoi3qns 61°€Q 3|qel eOLIBWY JO S31e1S pauun :ysn "4aqunu :N 'uswuiedw) sanu8od pliw (DA 'saxog Jo wns - Suiiey eipuswiag [ed1ul]) :9S-4a)d '3 uiroadodijody :Jody 'Bseasip s, Jawidyz|y :qv 6T '1C €9 SS°0- yoelg TEY "TEY 13 89°0- aUym VSN - ddey YSE '9G€ 1€ 85°0- Jlueds|H-uopN 00T '66 €TT €50 suedsiy vsn - Awduyla 0Z 'vz €9 ¢L O Jaelg LYT '8¥T 6T S€°0- uelsy SS9 'LL9 LC 6 o- aUYmMm 2oey STL'EVL T4 ov'0- slueds|H-uoN LOT '80T s g0- Jluedsiy Ay SzE '9T or Lo 5/3 89€ 'T8€ €¢ LEO- v/£-99 99T '8.1 9 80°0- 59> a8y S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y 19Q 98ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ €0C '€TC L1 6v0°0- adoun] YT 'OvT ¥ S0°0- eisy €TS '915 9T 800~ | eaBWY YyloON uoigay 192 'sL¢ € 2L0°0- Jaliied uoN 065 '009 0z 1¥0°0- Japue) snjejs 43 Jody (zeoz'le3e 0€€ 'T€€ £€C 190°0- av P MAQ uen) gy ALIYVID LT85 'vvS ST o0 DN dnou3qns |eduld TI¥ 'LOV w7 6€0°0- ON , . aujeseq ovv 83y €c 850°0- S$2A 1e uolledipaw qy snewoldwAs Jo asn qewaueda] Juimols syjuow 8T e SINODAV sdnousqns euL N 'oqa22e|d N % U] 92UdJBYIQ UedN paisnipy oeSdn0i3qng snolieA Aq syuow 8T Je SINOJAY AV ALIYV1D - eleq dnoidqns 0Z°€Q 3|qel *2JNSeaw awo31no Syl uo usawasocsdwi sjuasaldal anjea sanedau v, BOIS WY JO SIS PINUN 1YSN "Jaquinu :N 'Juduniedw) aAIMu80d pliw DN '3 uilosdodijody :Jody 'WBN-T 9BISqNS SAINUBS0D) - d|BIS JUBWSSASSY ISeS| S JPWIdYZ|Y :T80)-SyqQy 'Oseasip s awidyzly :av 6T 'T 60¢ 00°'T- Joeld 0EY '6tv €€ 59'1- 3UYM VSN - 9oeYy €S€ 'PSE 12 v€'1- | owedsig-uoN 00T '66 *CCC- 8T C- Jluedsiy vsn - Awduyia 0T 'vt 082 65°T- Yoelg 9vT 'S¥T S¢ oy' T- ueisy 759 VL9 8¢ 69'1- 3UYM soey TTL 'OvL 1¢ /€'T- | dwuedsiy-uon LOT '80T L8€ 90°'z- aluedsiH Apuyi3 ¥Z€ ''SIE o€ L9T- 6/2 S9€ '08€ 6¢ LV'T- ¥/-S9 S9T 'LLT T ¢6°0- S9> a8y €IV 'OV € L6'T- deiN Ty '29v 8T 86°0- ajewsa4 Xas GET 'TET €T €C0- 9108Az0WOH S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y 79qQ 98ed €20¢ 'M3lAaY d1WOU0D] pue [ed1ul]) JO) SINNISUIQ TZT '0€T €T €T elsv uoigay £LV 'ELY 23 6'T | eaLIaWY YUON . YET 'SPT Ly ELT 191489 UON sness 3 J0dy 6tS '1SS €€ Ul JaLuey (zzoz '|1 162 'L6 8€ 1L av PN dnoJ3qns (e ¥oAq uen) av ALI¥VID 6V '66¥ 8€ ¥9'1 DN o 6LE '€LE €€ €S'T ON aujaseq e YOb STV /€ 65 SOA uonedpaw qy 211ewoldwAas jo asn qewsueda' u-.__go_m syuow 8T 18 1av-1DIN N 'ogade|d N % -$2QV ul 2u3alayq uedsAl paisnipy sdnosdans o oeSdn0I3gns snoliep Aq syiuow 8T 1e 1aV-IDW-SIAV AV ALIMVID - eleq dnoi3qns TZ'€qQ 3|qeL B21J3WY JO SIRIS palunN ySN "agquinu :N 'Juauuiedw) sAiu8od pliw DI '3 uilosdodijody :3ody '91025 ausodwo) Qv SINODQAY 'eseasip s Jawisyzly :ay 6T 'T 0¢ 6T0°0- Joeld 0EY 'TEY 8¢ ¥S0°0- UYM VSN - 9dey €S€ '9G€ €C 2S00 dluedsiy-uoN 00T '66 €01 TS0'0- dluedsiH vsn - Ausiuyaa 0z 'v 0€ Tv0°0- joe|g LYT '87T 6T €390°0- uelsy YS9 '449 ec 0S0°0- IUYM 9oey VIL'svL T¢C 610'0- dluedsiH-uoN LOT '80T i 750°0- auedsiy Asiuyi3 Ve '91€ TE Q.00 QL3 L9€ 'I8€ S¢ 6v0'0- ¥L-99 99T '8.1 S 6000 S9> ady STV 'TTv 8¢ ¢80°0- 3leN v vov ot 7200- 3lewsaS X3S 9€T 'CET S- 6000 9108AzZO0WOH S '8ov Sz 950°'0- | 9108Az04919H 19T 'SLe € TL0°0- 1911Je2-UON snieys adAjouad 43 Jody S]U3IUO) JO 9|gE] 0} uImay €9Q @3ed 1oday a3uapiag |eul{ - aseasiq S, JawWidyz|y €207 'M3IASY J1WOUODT pue [edjul|] 10} SINIISUIQ "9JNSEIW SWOIIN0 SIYy} uo Juswdnoidwi sjuasaidal anjea annedau v, e21J3WY JO $31L1S PalUN (YSN '4aquinu :N 'quauuiedw] 9ARIUSOD pliw (DN '3 uiajoudodijody :30dy '(uoistdA [DA) Alojuaau| Sular] Ajleq Jo s31MAIDY - Apnis aAle1adoo?) aseasiq S dwWidyzly 1av-1DW-SOAV 'Oseasip s awidyzly :qy mfi. '0t «1CC- ST yoelg - S6€ 'C6E 174 €0 aYM SZ€E '6¢C€ ov 6z'Z | owedsiH-uopn A 06 98 8T€E ¢6'T aluedsiy LT 'T 81 80°'T Joeld LTT 'TET €T 9'1 ueisy asey ¥09 '919 oy 0c¢'c AUYM €59 'LL9 €€ 86T | owedsiH-uoN K 56 'S6 [{3 w0 suedsiy 687 VLT oy 8v'T Q=< £VE '9G€ 8¢ 720¢ ¥£-59 a3y TST '99T 9¢ 8C'T 99> ¥8¢€ '08¢ S LT'E e X35 66€ 9TV 6T T0'T ajewsaS L2T'TCT Sz €0°T 9108AzowoH e 'oY 9¢ /6T | 9108Az0J919H sniels adAjouas 3 Jody VET 'St LY €LC Ja1Ied-UoN 68T '€67 v Yard adoung S]U3IUO) JO 9|gE] 0} uImay Hoday a3uapIAg |eul{ - aseasiq S, JawWIdyz|y 9q a9ed €20 'M3IASY JIWOUOI3 PUE [EI1UID 10} AINUISUID juadJad 194 'Jagquinu :N 'saxog Jo wns - Supjey enuaWa( [EIUID (FS-HAD '9seasid s JaWIsyz|y Uo s|eu [ealul) :av.ld €100 662 £90°0- 658 'S/8 ale) |euosiag Z000°0 8'8¢ 860°0- 658 'S.8 $31qqOH pue aWoH 5000 71z L0°0- 658 'S/8 sileyy Aunwwo) (2auaisuod 100 9'€e €50°0- 658 'SL8 3UIA|0S W|q0Jd/uawWaspns av1d zzoz 112 33 uayod) av ALV ¥000°0 1'8¢ 180°0- 658 'SL8 UOREIUSLO 21000 5Lt LL0°0- 658 'S/8 Alows anjea d Suimols % 22ua1aHId qeisuessi sujewopqns gs-4dd euL uea|y paisnipy N 'ogadeld N »SUlBWOpQnNS g5-4aD AV ALYV - e1eq dnosdqns £2°€d 9|qel 9|gedijdde jou :yN '@insesw pajeadal 10j SIPPOW pPaxXIW INHININ 'SISAjeue 1ea11-03-1udlul paljipow 3] | jw 'SisAjeue jean) -03-JUSIU| 3] || '|EAISIUI SOUIPHUOI ([ "SaX0g JO WINS - Suley eljuawa( [edIUID :9S-4aD 'Uoisnya/ewapa 01 3np saliljewsouqge SuiSew| pajejaJ-pIojAwe 13-y |4y 100°0>d 6T-AIAOD 0 1dedw) 3y} ajen|eAs SYJUOIAI 8T 1€ 23uaiajjig luswieald] a8uey) uesa paisnipy 98uey) uean paisnipy "(9LT0- '€19°0- '1D %S6) ¥6E0- 80l €091 01 pajeadas WHININ Alewiid 6£7°0- '92L°0- 11D @mmm._womflM"m 6971 ¢/L9°T | -VIdV J3}je sjusuwissasse m:_._Omcww (zzoz @dudid3u0d . pajeadal WYWIN Asewnd avid) av ALIEVI 100°0>d 771 69T (1LL1) spalgns pazjwopuel '(vTT'0- 'v79°0- :1D %S6) YEV'O- [le uo paleadas WYIANIN 100°0>d VN VN L1|w ela paindwi (9L1°0- 'LEL'O- 1D %S6) 951°0- e1ep uissiw yum YAQDNY Juey anjen d (17 %s6) (qewauedd) syluow (oqade|d) syiuo 8T 1€ auljaseg wouy 8T 1€ sujjaseg wouy sasAjeuy AJARISUSS S-¥AD |elL gpSOSAjeuy ANAIMSURS 9S-HAD AV ALIMV1) - eleqg dnousqns zzZ'€q 3lqel S]U3IUO) JO 9|gE] 0} uImay G9Q @3ed 1oday a3uapiag |eul{ - aseasiq S, JawWidyz|y €707 'M3IASY J1WOUODT pue [edjul|] 10} SINISUIQ ol1el anjea ajeldn pazipiepuels tANS "QUSIDIHI0I uolssaI8ay 1y 'AydesSowo] uoissiwg uou}ISOd 113d 'ogaoe|d :0dd 'uediiusis jou :SN 'pariodal Jou YN 'YSI| JuswejyoInau (YN 'wesSo|y / wead)||iw :8%/8w 'DwodlnQ |ediul)) sausodwo) Ay :SINODAY 4N s0'0=d UN anjeaq (19897 UN 802°0- UN ¥ | Ienpialpul) sYIUOIA 8T 1€ OF/ZaV ewseld u] a8uey) "sA SINODAY Ul 38ueyd 4N SN 4N anjeA-d (19a37 uonendod) UN 90€°0- UN ¥ SYIUOIAI 8T 1€ OF/ZiaV ewiseld u] a8uey) "sA SINODAYV Ul 3dueyd 4N 9€0°0=d UN anjead (19A37 [enpiAIpuUl) SYIUOIA 8T 1€ JANS 13d Ul 38uey) "sA SINODIQAV ul a8uey) 4N 66T°0 4N ¥ UN 80'0=d UN anjea-d (19727 UN 7€8°0 UN Y uone|ndod) sYIUOIAl 8T € JANS 13d Ul d3ueyd *sA SINODAVY ul 33uey) suone|aLIo) 8€C zsT e N auljaseg oqgaoe|d Apjeamiq 8y/3w QT Alyuow 33/3w T suuy Apmis 0017C0Z AVLD "[e 13 Uosuems 924no0s 102-0009 = suoNe[2110) vZ'€d d|qeL S]U9IUO) JO 9|gE] 0} uIMay 1oday 93uapIA7 [eul{ - aseasiq S JaWIdYzZ|yY 99Q a8ed €70¢ 'M31A9Y dlWOou0I3] pue |edjul|) JO} 91NUISUIQ :uoisnpxy TTEL9LTOLON suojjedipaw qy panoidde ] syjuow p9 o3 dn JO 9SOp 3|qe)S JO dAIRU - *Syjuow uasoig g 1esi3 e USRI | peuiong r auyose cze3sA- | 09 o (amg o Koyl wouy a3ueyo pue TN AIFSWMW) | 8%/8w 0T) 3uswiean i~ (10z-0009 "GZ0T '0z Aenugay |, 310 "2d 'aa AQ uonaiduwios salyjewlouqe SuiSew; | Asowsw diposida ul Juawaiedw - Sunenui-ai asojaq 'A 'Gl 3se 11 9seyd) aseasiq vfimhsmw.Wc__omco 0} pajeja. piojAwe 4D 40 |13d (syuow g 98eiane I "qiI 95eUd s Jowipyzly Ales yum : : JO SJUDWISSISSE el aallsod plojAwe pawuyuod - | 'syluow g5-6) pouad s123lqns ul qewaued'd uojsua1xd 310 ; [HIN 'eiep A1ajes 3710 ejjuswap wswieaNn-Jo 110 Jo Aedy3 pue av pjiw Jo [DIA 01 anp av - 'Anqessjol 'Arajes uoisnpu| ?9jenjeag o1 Apnmis v aAlMsod AIH - 65989tV0.LON saJnpadoud Apnis yum aJaji91ul p|nod 1eyl swoldwAs uasoig g 1esiy Jo sisou8eip suleryoAsd 'sypene djwayasi jo Aoisy - (sv-€ avaHy) s - 91z 01dn Joj 0ad . 99M 9T¢ £¢< 3103s ISININ - S, JoWiayz|y jedtulpald LT0T 'ST 4990100 Jo ("Ajyuow Sx/8w | 1Dy 'GO 'Il 3seyd sam e uj aSueyd 0 940JS - Ajle3 yum suedpnaie A ' 13d Yy 04 Han 0T+ /3w § pue e YaMm d G20vd ul a8ueyd ¥3 Jody Aoamia Sx/Sw uj os|y pue plojAwy T 15e9)| 1B sassassod 'elpuswap & 1q /8w 0T + pajeAl|3 pue aseas|q 83j/8w 5) qewsueda yum pasouselp aAile[al S, JoWIdYz|y |ed1ulddld 92.35p 15414 Se yans s1o1oe) sy Yyum suedpiyed |[euollppe aAeY 1snW 9 01 GS - ul qewaue'd| sieaA 0g 01 g5 pase - YHM Judwieal | uoisnpu| Jo A1ajes pue Aesnya ajenjeag o) Apnis y gqewsaueda sa3eq uonajdwiod sawo0nQ Azewiid uonejndod juaned siojesedwo) usisaqg Apnmais Josuods |eu] / 9L pajewns3 sajpms Sujo3uQ '1'vd 3|qeL salpn3s Sulo3uQ 'va SJUaIUO) JO 9|geL 03 UIN1ay 1oday 27UapIAT [euUl{ - seasI( S JaWIsYyzZ|y £9Q 28eq €207 'M3IASY J1WOUODT pue [edjul|] 10} SINIISUIQ "ACE S[ELI|EJIUID) MMM :32IN0S "9JU3J3JU0I3|31 03PIA I A '|2qe| uado 10 '|ell pPajj0J1u0d paziwopued 1)y 'AydesSowo] uoissiwy U0JSOd :13d 'P3]|0J1u02-0gade|d :)d 'UoISudIXa |[agel-uado :370 '"BuiSew) SoueUOSAU J13BUSeW YN 'UoIlBUIWEXT 18IS [RIUSIAI-IUIA :ISININ 491Udd13nW DAl 'uoisualxa wJa3 8uoj 1317 'a|eas uoissatdaq JIeHIY :$OD 'WelSoIpJed043d3|3 (9D 'pullg 3|gnop :gq 'pPInj} [euldsouqalad 14SD 'WISAS SNOAJSU |BJIURD SND 'saxog Jo wng-3|eas Suney ejpuawaq [e21UlD :g9S-4add '9103 [eqo|D - Suizey elzuawaq [ea1ul)d :SO-4ad '¢3 udloidodijody 3 Jody 'aseasip s Jawisayz|y :av Aj@1eIndoE 51591 8ujwuopad juaned Jusaaid pjnod 1B} suoiipuoa [esipaw J3Y1Q - 503 BIA [BAIDIUI D] D/1D pa8uojold - av ueyi Jaylo sisouSe|p ejluUaWap E 31eJIpul pINO3J 1ey3 SUoIS3| JuediHUSBIS Ajjeaiun)d 134310 JO 3dUIPIAT - Suluueos 4 0} SUOIIBIIPUIBIIUDY) - 8221005 SO9 - sainpasoud Apnis yum 2194193u1 p|nod 1ey} swoldwAs Jo sisouSe|p su3e|ydAsd Auy - JedA 1 3sed ay3 ul saunzIds 10 '93041 "oeny J1wayos| judjsuel] Jo 3)0J3S e pey aABH - uawedw aaimudoo s, palgns ayl J0 9sned 8unnquuod e 3q ySiw ey} (av ueys Jayio) uopipuod |ea130joanau Jo |edipaw Auy - D5. Previous Systematic Reviews and Technology Assessments We identified one ongoing health technology assessment (HTA) conducted by NICE and two previously conducted systematic literature reviews (SLR). Both are summarized below. Additional SLRs were summarized in our previous report on aducanumab and are available in Section D7 of the 2021 review. NICE Gantenerumab for Treatment Early Alzheimer's Disease [ID 10668] As of June 2022, NICE is awaiting development to conduct an appraisal of the clinical and cost- effectiveness of gantenerumab for the treatment of MCl in early AD. Pang, M. et al. (2022). "Effect of Reduction in Brain Amyloid Levels on Change in Cognitive and Functional Decline in Randomized Clinical Trials: An Instrumental Variable Meta- Analysis">° Investigators reviewed the data of a 2021 variable meta-analysis'*® (cited in our 2021 report on aducanumab) that tested the amyloid hypothesis and resolved data quality issues in an update to ascertain the impact of the reduction in amyloid beta on clinical outcome measures in Alzheimer's disease. In addition to the 14 RCTs from the original publication, Pang et al. included data from the aducanumab-3 PRIME trial and the donanemab TRAILBLAZER-ALZ trial. The authors addressed several consistency issues from the original analysis regarding 12 of the 14 trials related to standard errors and misattributed data points. The included interventions and population characteristics have been summarized previously.?® The primary outcome of this analysis was the change in CDR-SB, ADAS-Cog, and MMSE, using the same representation of effect as the original meta-analysis. Sensitivity analyses were performed for 1) all published data, 2) only antibody data, and 3) only published antibody data. In the updated analysis, pooled results for "all data" showed statistically significant evidence of a causal relationship between a reduction in amyloid plaque and a reduction in cognitive and functional decline measured by CDR-SB (0.09; 95% Cl: 0.034- 0.15; p=0.0016). This was an improved result over the original meta-analysis, and the sensitivity analyses performed showed similar estimates. There were additional statistically significant causal effects for ADAS-Cog (0.33; 95% Cl: 0.12 - 0.55; p=0.0025) and MMSE (0.13; 95% Cl, 0.017- 0.24; p=0.024). While the meta-analysis builds upon Ackley et al. 2021 by including additional trials, there are limitations to this study. Similar to Ackley et al. 2021 meta-analysis, the Pang et al. meta-analysis faced limitations of data availability with many trials not reporting data on change in amyloid and change in a cognitive measure."? When data were available, it was predominately aggregated data. ©lnstitute for Clinical and Economic Review, 2023 Page D68 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Ackley & Glymour highlight the importance of using individual patient data which would allow the assessment of heterogeneity in response to amyloid reduction and the evaluation of relationship between amyloid reduction and cognition.? In addition, Pang et al. assumed no publication bias, non-informative loss to follow-up, consistency across radiotracers, and homogeneity in effects across drugs.>? Thus, there is a call for pharmaceutical companies to share individual patient data for investigators to more thoroughly examine the relationship between amyloid clearance and slowing of cognitive decline. Lacorte, E. et al. (2022). "Safety and Efficacy of Monoclonal Antibodies for Alzheimer's Disease: A Systematic Review and Meta-Analysis of Published and Unpublished Clinical Trials"*? Investigators identified all available registered trials for monoclonal antibodies (mAbs) evaluating mild cognitive impairment (MCl) due to Alzheimer's Disease (AD) at any stage, on two main registration databases: ClinicalTrials.gov (CT) and the European Clinical Trial Register (EUCT). Studies were excluded if they enrolled healthy participants, investigated drugs other than mAbs, or included any diagnosis other than MCl or AD. Results of the study were summarized narratively, and meta-analyses were conducted on frequency of adverse events (AEs) and severe adverse events (SAEs), frequency of ARIA-E and ARIA-H, mean change in CDR-SB score, and PET standardized uptake value ratio (SUVT). In total, 27 mAbs, assessed through 101 trials, were included in the review. Published and unpublished data were available for 12 of the 27 mAbs. Based on results from 17 studies on 7 mAbs, there was a significant effect on amyloid burden (as measured by SUVr) for patients treated with mAbs (Standardized mean difference: -0.88; 95% Cl - 1.30 to -0.47) compared to placebo. Results based on 16 studies on 8 mAbs concluded that patients treated with mAbs had a significantly less cognitive decline on CDR-SB {mean difference: -0.15). However, CDR-SB was mostly considered a secondary outcome, many trials were terminated due to futility, and the difference between groups did not reach clinical significance. The safety meta-analyses reported a higher frequency of AEs in patients treated with mAbs compared to placebo (Risk Ratio [RR]: 1.04; 95% Cl 1.02-1.06). However, no significant difference was shown between groups for SAEs (RR: 1.02; 95% Cl 0.96-1.09). Heterogeneity amongst the definitions and reporting of ARIA limited the ability to be able to compare the results across trials. To account for the heterogeneity, the data was stratified by mAb in an additional sensitivity analysis. ARIA-E had an RR of 10.65, with similar results shown in the sensitivity analysis (RR: 10.86). ARIA-H had an overall RR of 1.75, with a higher RR in the sensitivity analysis (RR: 2.11, 95% Cl 1.87-2.38). Ten RCTs did not report on ARIA-E or ARIA-H. Donanemab showed the highest risk for ARIA-E (RR: 34.63; 95% Cl 4.82-248.76) and ARIA-H (RR: 4.03; 95% Cl 2.09-7.79). There were some differences observed between ApoE €4 carriers (APOE+) and non-carriers (APOE-) for ARIA. Frequency of ARIA-E was slightly higher for carriers (RR: 13.47) than non-carriers (RR: 12.10), and a higher frequency of ARIA-H for non-carriers (RR: 2.18) than carriers (RR: 1.50). When stratified by ©lnstitute for Clinical and Economic Review, 2023 Page D69 Alzheimer's Disease - Final Evidence Report Return to Table of Contents mAb, APOE+ participants who received aducanumab had a higher frequency of ARIA-E (RR: 7.83 vs. RR: 2.96), as did APOE+ participant who received donanemab (RR: 30.32 vs. 8.51) or gantenerumab (RR: 41.61 vs. 3.02). ARIA-H frequency was higher for APOE+ participants who received aducanumab (RR: 3.05 vs. 1.85). Overall, the meta-analysis showed a significant effect of mAbs on amyloid burden, and the CDR-SB results were statistically but not clinically significant. A notable limitation to the analysis was the decision of the trial investigators to only present the clinical outcomes data as means and standard deviations, instead of using response rates or defining number of "responders". Thus, there is currently limited evidence to support that the removal of amyloid can have a substantial impact on cognition. In terms of safety, the meta-analysis concluded that patients have a higher risk of ARIA-E and ARIA-H when treated with mAbs compared to placebo. However, results are limited by the lack of standard definition of ARIA, most likely due to evolution of the definition of ARIA over time, and sensitivity analyses should be interpreted with caution. Villain, N. et al. (2022). "High-clearance Anti-amyloid Immunotherapies in Alzheimer's Disease. Part 1: Meta-analysis and Review of Efficacy and Safety Data, and Medico- economical Aspects">1120 Investigators reviewed the currently available evidence on the biological and clinical efficacy of high-clearance anti-amyloid immunotherapies in Alzheimer's disease. Using the data available from high-dose aducanumab (ENGAGE and EMERGE), donanemab (TRAILBLAZER-ALZ), high-dose lecanemab (G000-201 and CLARITY AD), and gantenerumab (GRADUATE 1 & 2), the investigators performed a meta-analysis on two cognitive measures (CDR-SB, ADAS-Cog). Data from CLARITY AD and GRADUATE 1 and 2 were only included in the CDR-SB and ARIA-E analyses. MMSE was not available for lecanemab. The meta-analysis results concluded that there was a significant effect of high clearance amyloid therapies on disease progression for CDR-SB (mean difference: -0.31; 95% Cl: -0.50, -0.11; p=0.002), ADAS-Cog (mean difference: -1.25; 95% Cl: -1.93, -0.57; p<0.001) but not for MMSE (mean difference: 0.31; 95% Cl:-0.19, 0.82; p=0.23). These results were confirmed with fixed effects and Bayesian analyses and when assessed based on disease severity {(MCl and mild AD) the improvements were maintained. While the results for CDR-SB support the theory that high- clearance anti-amyloid immunotherapies have a statistically significant effect after 18-27 months, the effect did not reach the established minimal clinically important difference (-1.63, -0.98). To address safety concerns, the investigators also performed a meta-analysis on ARIA. ARIA occurred at a significantly higher rate for participants on high-clearance anti-amyloid therapies, with the highest magnitude of the effect occurring for ARIA-E (RR = 10.98; 95% Cl: 7.06- 17.08; p<0.001). Despite in-trial monitoring and management, safety remains a concern. ©lnstitute for Clinical and Economic Review, 2023 Page D70 Alzheimer's Disease - Final Evidence Report Return to Table of Contents The investigators maintain that a meta-analysis should not replace the FDA's gold standard for approval, however, the results do confirm the trend regarding high-clearance anti-amyloid therapy's efficacy in removing amyloid. Additionally, the risk/benefit ratio for this class of drugs is still unknown and additional research needs to be done to identify responders. Longer follow-up data is also needed to improve their clinical relevance. ©lnstitute for Clinical and Economic Review, 2023 Page D71 Alzheimer's Disease - Final Evidence Report Return to Table of Contents E. Long-Term Cost-Effectiveness: Supplemental Information El. Detailed Methods Table E1. Impact Inventory Included in This Analysis | Notes on Sources (if Sector Type of Impact from [...] Perspective? quantified), Likely {Add additional domains, as relevant) Health Care . Magnitude & Impact Societal . Sector {if not) Formal Health Care Sector Health Longevity effects : : X X Health-related quality of life effects X X Outcomes Adverse events X X Paid by third-party payers X X . Paid by patients out-of-pocket X X Medical Costs Future related medical costs X X Future unrelated medical costs X X Informal Health Care Sector Pati i NA O Health-Related atler.\t time (.ZOStS - Unpaid caregiver-time costs NA X Costs - Transportation costs NA O Non-Health Care Sector Labor market earnings lost NA X Cost of unpaid lost productivity due to NA X Productivity illness Cost of uncompensated household NA O production Consumption Future consumption unrelated to health | NA O . . Cost of social services as part of NA O Sacial services . . intervention . Number of crimes related to NA O Legal/Criminal . . Justice intervention Cost of crimes related to intervention NA O . Impact of intervention on educational NA O Education . - achievement of population Housing Cost of home improvements, NA O remediation . Production of toxic waste pollution by NA O Environment . . intervention Other Other impacts (if relevant) NA O NA: not applicable Adapted from Sanders et al*?* ©lnstitute for Clinical and Economic Review, 2023 Page E1 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Target Population The model included a hypothetical cohort of individuals with MCI due to AD or mild AD entering the model and receiving either the intervention or comparator treatments. We assumed patients had amyloid positivity confirmed by a reliable method prior to initiating the model. In alighment with the clinical evidence, the starting population for the economic evaluation included adults with early AD, defined as MCI due to AD or mild AD. Consistent with population estimates, slightly more than half (55%) of the cohort started in the MCI due to AD health state, with the remaining cohort (45%) starting in the mild AD health state. Individuals could progress to more severe AD health states over the model time horizon. The majority of the cohort (92%) started the model in a community setting of care. Table E2 details the baseline patient characteristics for the model. Age influenced mortality and quality of life; sex influenced mortality. The baseline clinical stage and setting of care determined which health state and setting of care an individual started the model in. Table E2. Base-Case Model Cohort Characteristics Value Source Notes Mean Age, years 71 years Weighted average based on Female, % 52% van Dyck et al., 2022 sample size of each arm Clinical Stage, % AD population with MCI Due to AD 55% Potashman et al., 20202 underlying amyloid-beta Mild AD 45% pathology Setting of Care, % Percent of population ages Community 92% Johnson, 20192 65-74 who received long- Long-Term Care 8% term services and supports AD: Alzheimer's disease, MCI: mild cognitive impairment Treatment Strategies The list of interventions was developed with input from patient organizations, clinicians, and manufacturers on which treatments to include. Lecanemab was identified as an intervention to be included in the economic evaluation. At the initial stages of this assessment, donanemab was also included as an intervention, but was removed from the assessment following the complete response letter for accelerated approval. Lecanemab was evaluated in addition to supportive care, which could include non-pharmacologic and pharmacologic, but not disease-modifying, interventions. The comparator was supportive care alone. ©Ilnstitute for Clinical and Economic Review, 2023 Page E2 Alzheimer's Disease - Final Evidence Report Return to Table of Contents E2. Model Inputs and Assumptions Table E3 reports model assumptions along with their rationale that are important to consider when interpreting the findings. Table E3. Key Model Assumptions Assumption Rationale Lecanemab was effective at slowing the progression of disease while a patient had MCI due to AD or mild AD. Lecanemab was no longer effective once a patient reached moderate AD. The evidence on clinical outcomes that exists for lecanemab is in early AD. Based on trials of anti- amyloid therapies suggesting no benefit in moderate AD, clinical experts suggested there is likely no effect with anti-amyloid treatments at reducing disease progression once a patient has reached moderate AD. This assumption was tested in scenario analyses. Individuals stopped receiving lecanemab treatment once they reached moderate AD. Based on trials of anti-amyloid therapies suggesting no benefit in moderate AD, clinical experts suggested there is likely no effect with anti-amyloid treatments at reducing disease progression once a patient has reached moderate AD, and therefore in our model, treatment stopped once a patient reached moderate AD. In a scenario analysis, we modeled individuals stopping treatment once they reached severe AD. Robust evidence is lacking on lecanemab's effect on clinical outcomes after a patient has stopped the treatment, and thus no additional clinical benefit was assumed after a patient stopped treatment. All occurrences of ARIA and its associated consequences (on cost, quality of life, and treatment discontinuation) were modeled in the first year of treatment. ARIA has been observed as an adverse event for many studied treatments that target aggregated beta- amyloid. Consistent findings across these studies suggest ARIA occurs early in the treatment course. Caregiver impacts were incorporated in the societal perspective. The health care system perspective included the patient's cost and outcomes. Long-term care costs were incorporated in the health care system perspective. The health care system perspective included the cost and outcomes of the patient. Caregiver impacts were modeled as if each patient had one primary caregiver. Evidence on caregiver impacts was collected from a single, primary caregiver. AD: Alzheimer's disease, ARIA: amyloid-related imaging abnormality, CDR-SB: Clinical Dementia Rating-Sum of Boxes, MCI: mild cognitive impairment OInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page E3 Return to Table of Contents Model Inputs Model inputs were identified from best-available evidence and stakeholder engagement. The primary clinical inputs included the transition probabilities among alive health states, mortality, progressions to long-term care, treatment efficacy, the occurrence of adverse events, and discontinuation. Utility estimates were retrieved for both the patient and caregiver. The primary cost inputs included intervention acquisition costs, administration costs, monitoring costs, adverse event costs, long-term care costs, and other patient medical and pharmacy costs. Costs to inform the societal perspective included patient productivity, caregiver productivity, and caregiver health care costs. Clinical Inputs Transition Probabilities Between Alive Health States Table E4 provides the annual transition probabilities between each of the alive health states. These estimates were from a recent analysis of AD progression using data from beta-amyloid positive individuals in the National Alzheimer's Coordinating Center database.'?* Due to differences in age and sex (two characteristics that influence mortality) between the sample from the National Alzheimer's Coordinating Center and our baseline population characteristics described above, we calculated probabilities of transitioning to each health state conditioned on if an individual was alive. The calculation of these conditional probabilities normalizes the annual transition probabilities to be applied to our modeled population. The annual transition probabilities reported in Table E4. are the conditional probabilities and will be applied given the individual does not die in the model cycle. ©Ilnstitute for Clinical and Economic Review, 2023 Page E4 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Table E4. Transition Probabilities MCI Due to AD Mild AD Moderate AD Severe AD Source MCI Due to AD 77% 23% 0% 0% Mild AD 3% 58% 35% 4% Potashman et al., Moderate AD 0% 3% 55% 42% 2020124 Severe AD 0% 0% 2% 98% AD: Alzheimer's disease, MCl: mild cognitive impairment Mortality For each cycle, a risk of death was assigned based on age, sex, and health state occupancy. Age and sex-adjusted mortality served as the foundation for transitions to the dead health state, with an increased risk of death associated with AD dependent on the severity of AD. Age- and sex-adjusted mortality was sourced from US-specific life tables. Table E5. provides the relative risk of death from each health state. These relative risks were multiplied by the age- and sex-adjusted mortality for each model cycle. Table E5. Relative Risk of Death Based on Severity of Dementia Value Source Notes MCI Due to AD 1.82 Mild AD 2.92 125 Multiplied by age- and sex- Moderate AD 3.85 Andersen et al,, 2010 adjusted all-cause mortality Severe AD 9.52 AD: Alzheimer's disease, MCl: mild cognitive impairment Progressions to Long-Term Care Specific to each health state, the model tracked the setting of care (e.g., community or long-term care). Individuals with AD could progress from community to long-term care; however, once in long-term care, they remained there until death. Table E6. provides the annual probability of progressing to long-term care specific to each alive health state. These estimates are from an analysis that used Consortium to Establish a Registry for Alzheimer's Disease data.5! Table E6. Annual Transition Probabilities to Long-Term Care Value Source Calculated based on the reported mild AD annual transition MCI Due to AD 2.4% probability and the relationship between the relative risk of death for MCI due to AD and mild AD Mild AD 3.8% Moderate AD 11.0% Neumann et al., 199951 Severe AD 25.9% AD: Alzheimer's disease, MCI: mild cognitive impairment OInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page E5 Return to Table of Contents Treatment Effectiveness We assumed that lecanemab influenced disease progression from the MCl due to AD and mild AD health states. The evidence on clinical outcomes that exists for lecanemab is in early AD. Based on trials of anti-amyloid therapies suggesting no benefit in moderate AD, clinical experts suggested there is likely no effect with anti-amyloid treatments at reducing disease progression once a patient has reached moderate AD. We used best available evidence to estimate the effect of lecanemab on slowing disease progression from these health states. The most relevant clinical evidence for the model includes the rates of transitions among health states defined by AD severity. For lecanemab, Phase lll evidence existed on the progression to the next stage of dementia. That evidence served as the best available evidence to estimate the effect of these treatments on slowing disease progression from MCI due to AD and mild AD. Table E7. presents the estimates of treatment effectiveness we used in the model. These treatment effectiveness estimates were applied to the transition probabilities associated with disease progression reported in Table E4. Table E7. Treatment Effectiveness* on Slowing Progression Health State Lecanemab Notes MCI Due to AD 0.69 Equivalent to lecanemab's hazard ratio for slowing Mild AD 0.69 the progression of disease Moderate AD 1.00 Patient stopped treatment at moderate AD Evidence Source van Dyck et al., 20223% AD: Alzheimer's disease, CDR-SB: Clinical Dementia Rating-Sum of Boxes, MCI: mild cognitive impairment *Only applied to health state progressions (i.e., transitions to more severe health states). Adverse Events An important adverse event associated with beta-amyloid antibodies is the occurrence of amyloid- related imaging abnormalities (ARIA). ARIA typically occurs early in the treatment course and is often not associated with any symptoms. Table ES8. presents the probability of ARIA events for each treatment. We modeled that all ARIA occurred in the first model cycle. Later sections of this Supplement detail how the occurrence of these events influenced cost and quality of life. We did not model a risk of death from an ARIA event, but this is an important area for future research. Table E8. Adverse Events Parameter Lecanemab Probability of Any ARIA 21.5% Probability of Symptomatic ARIA 3.5% Source van Dyck et al., 202236 ARIA: amyloid-related imaging abnormalities ©lnstitute for Clinical and Economic Review, 2023 Page E6 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Discontinuation Evidence on discontinuation due to adverse events from the pivotal trials was used to estimate discontinuation over the first 12 months on treatment. We assumed individuals discontinued lecanemab due to adverse events halfway through the first model cycle (i.e., six months after starting treatment), in alighment with evidence that most serious adverse events occurred within the first six months of initiating the treatment. No discontinuation due to adverse events was assumed after the first year due to consistent findings that ARIA occurs at the beginning of the treatment course. Table E9. presents the treatment discontinuation rates due to adverse events reported in the pivotal trial. Table E9. Treatment Discontinuation Parameter Lecanemab AE-related Discontinuation of Treatment 6.9% Source van Dyck et al., 20223 AE: adverse event Separate from discontinuation due to adverse events, treatment stopped once a patient reached moderate AD. Stakeholders suggested there is likely no effect with anti-amyloid treatments at reducing disease progression once a patient has reached moderate AD, and therefore, treatment could discontinue. Discontinuation, either due to adverse event, disease severity or amyloid clearance, occurred halfway through the model cycle. Utility Inputs Health state utilities were derived from publicly available literature. These utility estimates primarily came from a cross-sectional study of patients with AD and caregivers with stratifications for both disease severity and setting of care.?® The utility weights were derived from the Health Utilities Index Mark Il (HUI:2) with weights based on the standard-gamble approach.'?® The HUI:2 is a commonly used instrument to calculate utility weights in the AD population because cognition is a separate attribute. The caregivers served as proxy respondents for the patient's quality of life, but also assessed their own quality of life.1?® Responses from the HUI:2 were converted to utility weights using the multi-attribute utility function developed for the HUI:2. We compared the utility estimates from this cross-sectional study to a recent systematic literature review published in 2020 and the estimates were comparable.'?" We elected not to select the recent systematic literature review estimates because the utility estimates were not stratified by care setting (e.g., community versus long-term care) and did not report quality-of-life estimates for the caregiver of the patient. Using the utility estimates from the recent systematic review would have required numerous assumptions and additional sources to be able to have utility estimates for individuals that live in the community, individuals that live in long-term care, caregivers of individuals that live in the community, and caregivers of individuals that live in long-term care. We understand the uncertainty ©lnstitute for Clinical and Economic Review, 2023 Page E7 Alzheimer's Disease - Final Evidence Report Return to Table of Contents around the utility estimates, and thus we varied each of these inputs across a wide range in sensitivity analyses. As shown in the results from our one-way sensitivity analysis, the range for each of our utility estimates for each level of disease severity includes the point estimate from the recent systematic review. The model used the utility estimates and the age of the people with AD from the cross-sectional study?® to calculate a disutility for each disease state and setting of care based off age-adjusted utility estimates. The calculated disutility was directly used in the model and was subtracted from age-adjusted utility estimates that varied based on age for each model cycle. Therefore, the model estimated quality of life was a function of age, disease severity, and setting of care. Table E10. presents the disutilities that were calculated from these estimates. Each disutility was applied for the duration of occupancy in the health state and setting of care. Table E10. Patient Disutility Estimates . . Long-Term Parameter Community Setting CaregSetting Source MCI Due to AD -0.17 -0.17 Calculated from utility Mild AD -0.22 -0.19 estimates and patient Moderate AD -0.36 -0.42 demographics in Neumann et Severe AD -0.53 -0.59 al., 199951126 AD: Alzheimer's disease, MCl: mild cognitive impairment In addition to the health state utilities reported in Table E10, a disutility of -0.14 was applied to people with AD experiencing symptomatic ARIA. This disutility was applied for the average duration of ARIA (12 weeks).102128 Thijs disutility estimate represents the disutility estimate for headache,'?° which was the most reported symptom among those with symptomatic ARIA.102:128 Impacts on the quality of life of caregivers was incorporated in the societal perspective. Caregiver utility estimates were calculated from the same cross-sectional study as the patient utility estimates described above.?® We used the age of the caregivers in the cross-sectional study*?® to calculate a disutility for each disease state and setting of care. The calculated disutility was directly used in the model. Importantly, the utility estimates reported in the cross-sectional study did not vary by AD disease severity (i.e., did not suggest a difference in caregiver utility for if the patient had mild, moderate, or severe AD). We adjusted these estimates to account for the difference in caregiver utility among AD disease severity reported in a study by Mesterton and colleagues.'*® The disutilities that were calculated from these estimates are presented in Table E11. The caregiver disutility was applied onto the patient's utility estimate. No caregiver disutility was assigned upon or following the patient's death. ©lnstitute for Clinical and Economic Review, 2023 Page E8 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Table E11. Caregiver Utility Estimates Parameter Caregiver Disutility Source MCI Due to AD -0.03 Calculatetil from utility est_lma_\tes and patient demographics in . Neumann et al., 1999;51.126 Mild AD 0.05 adjusted for AD severity using relationship from Mesterton et al., Moderate AD -0.08 130 2010 Severe AD -0.10 AD: Alzheimer's disease, MCl: mild cognitive impairment Economic Inputs Drug Utilization The following inputs were used to model drug utilization and associated costs: Route of administration Dosing Frequency of administration Duration of treatment Table E12 reports these characteristics for lecanemab. Table E12. Treatment Regimen Recommended Dosage Generic Name Lecanemab Manufacturer Eisai Co., Ltd Route of Administration Intravenous Dosing 10 mg/kg Frequency of Administration Every 2 weeks Duration of Treatment Until moderate AD Source van Dyck et al., 20223 AD: Alzheimer's disease, kg: kilogram, MCI: mild cognitive impairment, mg: milligram Intervention Costs Given that a net cost is not yet available for lecanemab, we used the wholesale acquisition cost as the price for lecanemab. The annual cost in Table E13 does not include any provider-administered mark-up (assumed to be 6% in addition to the cost in Table E13), or any treatment-associated administration or monitoring costs. ©Institute for Clinical and Economic Review, 2023 Page E9 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Table E13. Drug Acquisition Costs Dru Annual Wholesale Source g Acquisition Cost* Lecanemab $26,500 REDBOOK?3! *Doesn't include any provider-administered mark-up (assumed to be 6% in addition to the cost in table E13). Non-Intervention Costs Costs outside of drug acquisition are stratified by perspective below. Health Care System Costs Administration Costs Lecanemab is administered by way of intravenous administration. We assumed an average administration cost of $78.35 per administration (HCPCS code 96365).132 Monitoring Costs For the first year while a patient used lecanemab, we assumed they were monitored for ARIA using brain magnetic resonance imaging (MRI) every three months. Because evidence suggests the vast majority of ARIA occurs within the first year of treatment, no MRIs were modeled after the first year on treatment. We assumed an average brain MRI cost of $261.10 per scan (HCPCS code 70553).132 Adverse Event Costs In addition to the brain MRIs described above for monitoring, if a patient experienced an ARIA event, the patient received a brain MRI every four weeks until the ARIA was either resolved or stabilized."> The average duration of an ARIA event was 12 weeks; therefore, a patient that experienced an ARIA event received three additional brain MRIs associated with managing the adverse event. We assumed an average brain MRI cost of $261.10 per scan (HCPCS code 70553).13? Non-Treatment Related Health Care Costs Annual medical costs stratified by disease severity were sourced from a study conducted by Leibson and colleagues.'3 This study reported the average annual inpatient and outpatient medical costs for people who were cognitively normal, had MCl, were newly diagnosed with dementia, and had prevalent dementia. We assumed costs associated with the newly diagnosed dementia group corresponded to the mild AD health state, and costs associated with the prevalent dementia group corresponded to the moderate and severe AD health states. We assumed the annual medical costs were the same for people with AD in the community or in long-term care. Using these estimates, we calculated a cost multiplier for each health state in the model based on those that were cognitively normal. In the model, we multiplied this cost multiplier by the average age-adjusted ©lnstitute for Clinical and Economic Review, 2023 Page E10 Alzheimer's Disease - Final Evidence Report Return to Table of Contents health care costs for the US general population. These annual costs were included in the model to account for related and unrelated medical health care utilization, stratified by disease severity. Table E14. reports these cost multipliers that were applied to the health care costs of the general population. Table E14, Direct Medical Cost Multipliers Health State Multiplier Source MCI Due to AD 1.12 Mild AD 1.56 . . Moderate AD 1.93 Leibson et al., 2015 Severe AD 1.93 AD: Alzheimer's disease, MCl: mild cognitive impairment To capture other pharmacy costs not related to lecanemab, we assumed 33.3% of individuals with mild AD received generic donepezil 10 mg once daily ($0.21 per day)*3! and 33.3% of individuals with moderate AD received generic memantine 10 mg twice daily ($0.66 per day).13%:134 Long-Term Care Costs For people with AD in the long-term care setting, additional costs associated with long-term care were included. Table E15. lists the monthly costs for long-term care that were assigned to those individuals who progressed to the long-term care setting. Table E15. Long-Term Care Costs Parameter Value Source Notes Administration on Skilled nursing facility Aging'® cost Costs have been inflated from 2016 US dollars to 2021 US dollars using the price index for health care services. Long-Term Care $7,394 per month 136 Societal Costs Patient Productivity Costs A study published in 2020 by Robinson and colleagues reported that among people with amyloid- beta positive MCl, 20.4% reported still working, with 4.9% of those who worked reporting a reduction in work due to AD.'3" Similarly, among people with beta-amyloid positive mild AD, 11.2% reported still working, with 8.6% of those who worked reporting a reduction in work due to AD.>¥' We assumed 0% of individuals with moderate and severe AD work with the reason for non- employment not attributed to AD. The average age of the population in the Robinson study was comparable to the average age of our modeled cohort. For those individuals who reduced work due to AD, we assigned lost productivity costs of 20 hours per week. The average hourly wage of $32.46 was used to monetize the lost productivity.1®® ©lnstitute for Clinical and Economic Review, 2023 Page E11 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Caregiver Productivity Costs The Robinson et al., 2020 study also reported caregiver time spent caregiving for individuals with MCI.237 A separate source by Haro and colleagues reported caregiver time spent caregiving for community-dwelling people with mild, moderate, and severe AD.23*® Table E16. reports the average caregiver time spent caregiving for community-dwelling people with AD in each health state that were used in the model for people with AD dwelling in the community. Time includes time spent providing supervision and activities of daily living (basic and instrumental). Table E16. Caregiver Time Spent Caregiving for Community-Dwelling Caregivers Health State Value Source MCI Due to AD 69 hours/month Robinson et al., 202037 Mild AD 113 hours/month Moderate AD 169 hours/month Haro et al., 2014'* Severe AD 298 hours/month AD: Alzheimer's disease, MCl: mild cognitive impairment The What Matters Most study, sponsored by the Alzheimer's Disease Patient and Caregiver Engagement consortium, suggested caregiver time spent with long-term-care-dwelling people with AD was 44% that of caregiver time spent with community-dwelling people with AD; and thus the estimates reported were multiplied by 44% to estimate the caregiver time spent for long-term-care- dwelling people with AD.* The average hourly wage of $32.46 was used to monetize the time spent caregiving.'3® Caregiver Direct Medical Costs Table E17. presents the direct medical costs for the primary caregiver of a patient with AD. These are the same values we used in our prior AD review but have been inflated to 2021 US dollars. Table E17. Caregiver Direct Medical Costs Health State Value Source MCI Due to AD $460 per month . 137 Mild AD $965 per month Robinson et al, 2020 Moderate AD $1,544 per month Robinson et al, 2020'%7 & Mesterton et Severe AD $1,930 per month al., 2010130 AD: Alzheimer's disease, MCI: mild cognitive impairment OInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Page E12 Return to Table of Contents E3. Results Description of evLY Calculations The evlY considers any extension of life at the same "weight" no matter what treatment is being evaluated. Below are the stepwise calculations used to derive the evlY. 1. First, we attribute a utility of 0.851, the age- and gender-adjusted utility of the general population in the US that are considered healthy. 14 2. For each cycle (Cycle ) in the model where using the intervention results in additional years of life gained, we multiply this general population utility with the additional life years gained (ALYG). 3. We sum the product of the life years and average utility {cumulative LYs/cumulative QALYs) for Cycle | in the comparator arm with the value derived in Step 2 to derive the equal value of life years (evLY) for that cycle. 4. |If no life years were gained using the intervention versus the comparator, we use the conventional utility estimate for that Cycle I. 5. The total evlLY is then calculated as the cumulative sum of QALYs gained using the above calculations for each arm. 6. We use the same calculations in the comparator arm to derive its evlLY. Finally, the evLYG is the incremental difference in evLY between the intervention and the comparator arms. E4. Sensitivity Analyses Table E18. Credible Intervals from Probabilistic Sensitivity Analysis for Lecanemab versus Supportive Care Model Outcome Lecanemab Supportive Care Total Costs $492,000 $364,000 Total QALYs 3.87 (3.42, 4.40) 3.35(3.08, 3.61) ICER ($/QALY) $248,000 Total evlYs 3.99 (3.47, 4.57) | 3.35(3.08, 3.61) ICER ($/evLY) $200,000 evlY: equal-value life year, QALY: quality-adjusted life year ©lnstitute for Clinical and Economic Review, 2023 Page E13 Alzheimer's Disease - Final Evidence Report Return to Table of Contents E5. Scenario Analyses We conducted numerous scenario analyses to test the structural assumptions that were made. Scenario Analysis 1: Treatment Stop at Severe AD The incremental cost-effectiveness ratios are provided in the main report, but we include the model outcomes in Table E19. and Table E20. Table E19. Model Outcomes for the Health Care Sector Perspective, Treatment Stop at Severe AD Intervention . Years in the Treatment Cost* Total Cost | Life Years QALYs evLYs Community Lecanemab $137,000 $529,000 | 6.34 3.93 4.08 4.28 Supportive Care SO $363,000 | 5.77 3.34 3.34 3.69 evlLYs: equal-value life years, QALYs: quality-adjusted life years *Intervention cost doesn't include provider administered mark-up {(modeled as 6%), monitoring costs, or administration costs. Table E20. Model Outcomes for the Modified Societal Perspective, Treatment Stop at Severe AD Intervention . Years in the Treatment Cost* Total Cost | Life Years QALYs evlLYs Community Lecanemab $137,000 $833,000 | 6.34 3.57 3.76 4.28 Supportive Care $0 $670,000 | 5.77 2.98 2.98 3.69 evlLYs: equal-value life years, QALYs: quality-adjusted life years *Intervention cost doesn't include provider administered mark-up {(modeled as 6%), monitoring costs, or administration costs. Scenario Analysis 2: Updated Caregiver Disutility Estimates In this scenario analysis, we updated the source for the caregiver disutility estimates to estimates reported in a recent conference poster.1#? Table E21 reports the caregiver disutilities, stratified by patient health state and setting of care, that were applied in this scenario analysis. These disutilities were not used in our base-case analysis for multiple reasons including the general appropriateness of using time tradeoff for eliciting utilities for a caregiver, the challenge for a person from the general population to detangle the patient health states from the caregiver quality of life in this exercise, the uncertainty in how the domains in the study mapped to our health states and setting of care, the limited information available regarding the framing of the questions, and the small sample size. Page E14 Return to Table of Contents OInstitute for Clinical and Economic Review, 2023 Alzheimer's Disease - Final Evidence Report Table E21. Caregiver Disutilities for Scenario Analysis Health State Community Setting Long-term Care Setting MCI due to AD -0.04 -0.04 Mild AD -0.06 -0.11 Moderate AD -0.20 -0.14 Severe AD -0.36 -0.21 AD: Alzheimer's disease, MCl: mild cognitive impairment These disutilities are larger than the estimates used in our base-case, and thus generated lower incremental cost-effectiveness ratios from the societal perspective. Table E22. reports the updated societal perspective cost-effectiveness estimates assuming these disutilities. Table E22, Incremental Cost-Effectiveness Ratios, Updated Caregiver Disutility Estimates Lecanemab vs. Supportive Care Perspective Cost per Life Year Gained Cost per QALY Gained Cost per evLY Gained Societal $265,000 $213,000 $163,000 evlY: equal-value life year, QALY: quality-adjusted life year E6. Model Validation Model validation followed standard practices in the field. We used several approaches to validate the model. First, we provided the preliminary model structure, methods and assumptions to manufacturers, patient groups, and clinical experts. Based on feedback from these groups, we refined data inputs used in the model, as needed. Second, we varied model input parameters to evaluate face validity of changes in results. We performed model verification for model calculations using internal reviewers. As part of ICER's efforts in acknowledging modeling transparency, we also shared the model with the relevant manufacturers for external verification shortly after publishing the draft Evidence Report. Finally, we compared results to other cost-effectiveness models in this clinical area. Prior Economic Models We compared our findings for lecanemab to a peer-reviewed publication on the potential economic value of lecanemab.'** The peer-reviewed publication used a patient level simulation model (AD ACE) with the assumption that the treatment effect was driven by reductions in amyloid PET levels using evidence from the Phase I trial.2** In our model, we used a Markov model with the treatment effect modeled as the hazard ratio for observed progressions to the next stage of dementia using evidence from the Phase lll trial. Despite the difference in structure (Markov model versus patient level simulation) and treatment effectiveness estimates (hazard ratio on progressions to next stage of dementia versus amyloid PET level), our findings are relatively similar. The peer-reviewed publication reported a threshold range {from $50,000 to $200,000 per QALY gained) of $9,000 to $38,000.1*3 Our reported threshold range (from $50,000 to $200,000) was $3,000 to $29,000. ©Institute for Clinical and Economic Review, 2023 Page E15 Alzheimer's Disease - Final Evidence Report Return to Table of Contents Differences in the findings can largely be explained by the fact that our model used direct clinical outcome data from the Phase lll trial whereas the peer-reviewed publication used amyloid PET level as a surrogate predictor of clinical outcomes. There is also a peer-reviewed publication on the cost-effectiveness of a hypothetical disease- modifying treatment for AD that modeled various treatment strategies, including a continuous dosing strategy and a fixed duration strategy.'** The peer-reviewed publication used model inputs for a hypothetical intervention and reported an incremental cost-effectiveness ratio of $612,000 per QALY gained under a continuous treatment strategy and an incremental cost-effectiveness ratio of $126,000 per QALY gained under a fixed duration strategy where 40% discontinued at 6 months and 100% discontinued at 18 months.»** Our model nearly replicated those incremental cost- effectiveness ratios when we made similar model assumptions (e.g., treatment effect, treatment duration, discontinuation, treatment cost). When we updated our model to reflect the inputs used for the hypothetical treatment that was modeled in this peer-reviewed publication, our model generated an incremental cost-effectiveness ratio of $625,000 per QALY gained under a continuous treatment strategy and an incremental cost-effectiveness ratio of $143,000 per QALY gained under a fixed duration strategy. In our prior AD review that included aducanumab, we calculated an optimistic treatment benefit scenario that closely reflects the assumptions we have made in our base-case analysis of this review.? In the optimistic treatment benefit scenario of our prior AD review, we assumed the hazard ratio for the trial that showed a benefit (i.e., we did not blend the hazard ratio with the trial that did not show a benefit) and we assumed the same treatment effect for transitions out of MCl and mild. The threshold range, using thresholds of $100,000 to $150,000, for aducanumab based on the optimistic treatment benefit scenarios was $11,000 to $25,000. This is similar to the threshold range we report for lecanemab. The threshold range, using thresholds of $100,000 to $150,000, for lecanemab was $8,900 to $21,500. This is slightly lower than the range for aducanumab that assumed the optimistic treatment benefit due to lower discontinuation for lecanemab (which increases treatment costs for the cohort), a higher baseline starting age for lecanemab, and differences in how discontinuation was programmed. ©lnstitute for Clinical and Economic Review, 2023 Page E16 Alzheimer's Disease - Final Evidence Report Return to Table of Contents F. Supplemental Policy Recommendations Payers Coverage Criteria: General ICER has previously described general criteria for fair coverage policies that should be considered as cornerstones of any drug coverage policy: see Cornerstones of "Fair" Drug Coverage: Appropriate Cost-Sharing and Utilization Management Policies for Pharmaceuticals. Drug-Specific Coverage Criteria: Lecanemab The large number of patients with MCl and mild dementia due to AD, combined with the potential for side effects and the high annual price for lecanemab, will lead payers to develop prior authorization criteria and to consider other limits on utilization. None of these limits, however, should undermine the tenets of fair access to which all patients have a fundamental right.1** To explore the appropriate application of evidence to coverage policy, and to reflect the views of patient experts and clinicians on specific ways that payers might appropriately use coverage policy to manage resources prudently, we present the following perspectives on specific elements of cost sharing and coverage criteria for lecanemab. Coverage Criteria e Age: Age criteria are likely to follow the age range from the clinical trials for lecanemab, which encompass ages 50-90 years old. However, it is not unreasonable for payers to consider excluding patients <65 years old from coverage. Subgroup analyses from the pivotal trial are exploratory and may be confounded by ApoE €4 status, but there was no statistically significant benefit from treatment among patients <65 years old. In considering this age cutoff for coverage, payers should carefully weigh the evidence in light of the likelihood that patients under age 65 and their clinicians may not view the subgroup analysis as persuasive. e Clinical eligibility: Because of its modest efficacy and potential for harm, payers are likely to closely adhere to clinical trial inclusion criteria. Treatment with lecanemab is indicated in patients with MCI due to AD and mild AD, with biomarker evidence of Alzheimer's disease. o Diagnosis of MCI and mild dementia should be based on cognitive and functional assessments of patients. Some assessments used in the CLARITY-AD clinical trial (e.g., Weschler Memory Scale IV-Logical Memory (subscale) 11} are not commonly used in clinical practice. Thus, it is reasonable to define MCl and mild AD with more ©Ilnstitute for Clinical and Economic Review, 2023 Page F1 Alzheimer's Disease - Final Evidence Report Return to Table of Contents commonly used, validated scales such as the MMSE, CDR-GS, St. Louis University Mental Status (SLUMS) scale, or the Montreal Cognitive Assessment (MoCA). For example, the VA has established the following criteria: MMSE >21, SLUMS or MoCA >16, and Functional Assessment Staging Test (FAST) Stage score or 2-4. o Determination of AD versus other causes of dementia: To exclude other causes of dementia, payers are likely to require a screening MRI within the previous year that does not show evidence of acute or sub-acute hemorrhage or diffuse white matter disease. Although tests to demonstrate the presence of amyloid will be the next step in insurance coverage for most payers, some may also request that blood tests be done to exclude other causes of dementia, including tests for syphilis, thyroid disease, and vitamin B12 deficiency, and screening for depression. To establish amyloid presence in the brain, payers will have the choice of covering PET scans and/or CSF-based testing and should cover both to provide broad options for patients and clinicians in different practice settings. o Testing for ApoE €4: Subgroup analysis of CLARITY-AD data suggests that there may be lower efficacy of lecanemab in ApoE €4 homozygotes, which may represent the higher risk of ARIA in this group. Given the modest efficacy of lecanemab and potentially severe consequences of ARIA, it is not unreasonable to require testing for ApoE €4 status prior to treatment. However, payers should be aware that some patients may be reluctant to be tested due to ethical implications of testing positive for the mutation (e.g., impact of knowledge of mutation on children). Additionally, if payers require ApoE testing, it may be more efficient to do ApoE testing prior to a PET scan or CSF testing for amyloid, as patients who are homozygous for the mutation may not be candidates for treatment regardless of amyloid status. e Exclusion criteria: Exclusion criteria in the CLARITY-AD clinical trial focused on patients who had evidence of dementia caused by etiologies other than AD and patients who may be at higher risk for complications from treatment. o Evidence of any medical, neurological, or mental health condition that may be a contributing or primary cause of cognitive impairment. For example, significant lesions on brain MRI that indicate another cause of dementia; untreated thyroid disease, vitamin B12 deficiency, or HIV; transient ischemic attack, stroke, or seizures; brain tumors. o Patients at increased risk of cerebral bleeding (e.g., evidence of cerebral micro- or macrohemorrhages, superficial siderosis, vasogenic edema, aneurysms or vascular malformations, lacunar infarcts or severe small vessel or white matter disease on ©lnstitute for Clinical and Economic Review, 2023 Page F2 Alzheimer's Disease - Final Evidence Report Return to Table of Contents screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or screening MRI, uncontrolled bleeding disorder, platelet count <50,000 or