Romano (DeMuro) C, Mogle J, Gold A, Hartry A, Taylor J, Callahan LF, Monks D, Kremer I, Lappin D, Frangiosa T, Lee J, Shirneshan E, Slowiejko D, Paulsen R. Best-fit clinical outcomes assessment measures to evaluate what matters most in Alzheimer's disease: identifying a core outcomes set from existing measures. Poster presented at the 18th Clinical Trials on Alzheimer's Disease (CTAD) 2025; December 4, 2025. San Diego, CA.

BACKGROUND: The What Matters Most (WMM) in Alzheimer’s disease (AD) research program seeks to identify and measure disease- and treatment-related needs, preferences, and priorities of people at risk for or living with Alzheimer’s disease (PLWAD) and their care partners. This program includes an ongoing series of studies sponsored by the Alzheimer’s Disease Patient and Caregiver Engagement(AD PACE) Initiative of UsAgainstAlzheimer’s. Previous WMM research identified concepts encompassing symptoms, functioning, and outcomes that were endorsed as important by individuals at risk for or living with mild cognitive impairment (MCI) or mild AD and care partners of those with moderate or severe AD living in the United States (US). This information was used to develop a conceptual model of disease, comprising 50 concepts across 6 domains: Thought Processing, DailyActivities, Emotions, Communication, Socia lLife/activities, and General Independence. Subsequently, an observational survey-based study was conducted to assess some psychometric properties of the conceptual model and to quantify the prioritization and impact of WMM concepts on daily life among a diverse population of PLWAD and care partners representative of the full spectrum of AD severity. To advance the measurement of concepts that matter most to individuals at risk for or living with AD, this study compared WMM concepts to existing clinical outcome assessments (COAs) including selected neurocognitive measures commonly used in AD clinical trials with a goal of identifying a core set of “ best-fit” COAs. Candidate COAs were considered those that best balanced (1)broad coverage and rigorous measurement of the WMM concepts with (2) use or potential use in clinical trials.  Results can guide AD researchers in the selection of patient-centric outcome measures with a demonstrated link to priorities among those living with cognitive impairment and their care partners. Further, the concept mapping presented here also provides insights into the limits of existing measures relative to WMM priority concepts, offering a foundation for future support of regulatory and reimbursement evaluations of new drugs.

METHODS: A targeted literature review was conducted to identify candidate COAs, meaning measures with appropriate development and established measurement properties specific for use across the spectrum of AD severity that may capture the WMM concepts as identified by the AD PACE Initiative. ClinicalTrials.gov was searched to identify COAs, including neurocognitive measures that have been used in registries, interventional studies, and recent drug approvals in the US within the past 5 years(2012-2022 primary completion dates). A subset of candidate measures was selected for concept mapping based on relationship to atleast 1 WMM domain, measurement properties, and utility (e.g., number of items, response window) of the measure. Consistent with previous mapping of widely used COA items to WMM concepts, items from both frequently used COAs and newly identified, additional(“ supplemental”) COAs were reviewed and matched either directly or indirectly to WMM concepts. Evidence for each candidate COA was reviewed to determine how existing measures meet needs for assessing the WMM concepts that are of greatest importance to individuals at risk for or living with AD or their care partners. Measurement gaps, where WMM concepts lack a corresponding measure, were identified.

RESULTS: Among 20 COAs identified as most frequently used in clinical trials, 8 were selected for mapping: ADCOMS, ADCS-ADL-MCI, CDR (including CDR Sum of Boxes),FAQ, iADRS, NPI/NPI-Q, PACC, and QoL-AD. Searches for additional (“supplemental”)measures for the WMM model identified an additional 20 measures, 11 of which were selected for mapping of comprehensive coverage of WMM concepts. These included the A-IADLs, ADRQL, AES, BADLs, CMAI,CFIa, CogState C3, Daily Diary Assessments, eCog, NIH Toolbox, and PROMIS/Neuro-QoLShort Forms. Subject-, care partner-, and clinician-reported versions of tools were included. As expected, no single measure was  identified that provided complete, direct coverage of all WMM concepts. The selected best-fit measures offered the broadest direct orindirect alignment with the WMM model. However, important measurement gaps were identified regarding coverage and assessment of specific concepts/domains relevant to the WMM prioritization and impact evaluations.

CONCLUSION: This work identifies existing COAs that align with concepts from among those identified as “ What Matters Most” to PLWAD and their care partners. This work also provides guidance for selecting appropriate outcome measures for the intended population, targeted concepts, and hypothesized trajectory of change relative to the patient-centric WMM concepts. We propose 2 best practice recommendations for implementing measurement of patient-centric concepts: (1)specifically identify and select a combination of measures that provide robust coverage of the concept(s) of interest, and (2) ensure selected measures provide direct coverage of target WMM concepts from multiple perspectives whenever possible (i.e., PLWAD, care partner, clinician). Critically, this work demonstrates the unmet need and measurement gap for WMM concepts that remain indirectly or incompletely covered by existing COAs.