Developing a microsimulation model of Alzheimer's disease for the evaluation of disease modifying therapies

Clinical trials of amyloid-targeting therapies (ATTs) for Alzheimer's Disease (AD) showed slowing of decline in continuous cognitive and functional outcomes over eighteen months of follow-up. Existing simulation models used to extrapolate these results do not account for the distinct but correlated trajectories of cognitive and functional decline, limiting their accuracy in evaluating ATT effects on both domains. We developed an AD microsimulation model using an amyloid-positive memory clinic population that can translate slowed cognitive decline into effects on clinical functional independence stages. The model contained stages of mild cognitive impairment (MCI), community dementia, institutionalisation, and death. The transitions from MCI to dementia and from dementia to institutionalisation or death were related to Mini-Mental State Examination (MMSE) measurements through statistical joint models. Model validation showed that observed and predicted cumulative incidence curves overlapped for all stage transitions, indicating an accurate description of time in each stage. An untreated cohort starting with MCI had a mean duration of MCI, mild, moderate and severe community dementia, and institutionalisation of 3.3, 2.3, 1.6, 1.1, and 1.7 years, respectively, totalling 10.0 years. When ATTs slowed MMSE decline by 30% during MCI and mild dementia (community dementia with an MMSE above 20), the duration of MCI and mild dementia was prolonged by 0.4 and 1.0 years, respectively, and institutionalisation was delayed by 1.1 year. The time institutionalised remained similar. Our AD microsimulation model can translate ATT effects on cognition into effects on functional independence and can be used to assess the cost-effectiveness of new therapies.