(B-145) Timing of Symptomatic Treatment Discontinuation on Neuropsychiatric and Functional Outcomes in Alzheimer's Disease and Related Dementias

Background: Symptomatic treatments—acetylcholinesterase inhibitors (AChEIs) and memantine—are widely used in Alzheimer’s disease and related dementias (ADRD). Due to their uncertain long-term effects, deprescribing is recommended to reduce drug burden and the potential for adverse events. However, the optimal timing for discontinuing symptomatic treatment is unknown.

Objectives: To evaluate the impact of different timings of symptomatic treatment discontinuation on neuropsychiatric and functional outcomes within two years of treatment initiation in individuals with ADRD.

Methods: A retrospective cohort study of individuals aged ≥65 with baseline mild dementia was conducted using Merative® MarketScan with Medicare supplemental databases and a target trial emulation approach. Study cohort consisted of patients who initiated symptomatic ADRD treatment between 2011 and 2020. Neuropsychiatric outcome was assessed through antipsychotic initiation; functional outcomes included the incidence of falls or related injuries and progression to moderate-to-severe dementia. Four treatment strategies compared were “discontinue treatment after continuous treatment for m months, where m is < 4 [strategy 1], 4–6 [strategy 2, reference group], 7–12 [strategy 3], or 13–18 [strategy 4].” We used a weighting approach to mitigate baseline and time-varying confounding and selection bias.

Results: Separate cohorts were analyzed for each respective outcome of interest: 74,481 patients with mild dementia at baseline, 39,288 patients with mild dementia but no baseline antipsychotic use, and 48,611 patients with mild dementia but no baseline falls or related injuries. The risk of progression to moderate-to-severe dementia was similar across treatment strategies, with a 2-year risk between 28.7% and 29.7%. Strategy 2 had the lowest 2-year risk of antipsychotic initiation (11.1%). Compared to strategy 2, the risk differences for antipsychotic initiation at 2 years were 2.1% (95% CI: 0.3%, 3.6%) for strategy 1, 1.8% (95% CI: 0.2%, 3.6%) for strategy 3, and 4.1% (95% CI: 2.6%, 5.6%) for strategy 4. For falls or related injuries, strategies 2, 3, and 4 had a similar 2-year risk, ranging from 25.1% to 26.6%. Strategy 1 had a higher risk, with a risk difference of 2.2% (95% CI: 0.1%, 4.7%) compared to strategy 2.

Conclusions: The timing of discontinuing symptomatic ADRD treatments influences clinical outcomes. Our findings suggest that discontinuing after continuous treatment for 4–6 months may provide better long-term effects in neuropsychiatric and functional domains. However, determining whether to discontinue treatment requires a comprehensive evaluation and should be tailored to each patient's specific needs.