Abstract
Introduction: In Alzheimer’s disease (AD), significant behavioral symptoms are associated with patient distress, caregiver burden, admission to long-term care facilities, and use of psychotropic medications. Because of safety risks involved with antipsychotic use in this patient population, evaluating the efficacy of antidementia drugs on behavioral symptoms is of great interest. A new, extended-release (ER) formulation of memantine (28 mg, once daily) has been approved in the US, based upon the results of a 24-week, multinational, randomized, placebo-controlled trial (MEM-MD-50, NCT00322153) in patients with moderate to severe AD concurrently taking a cholinesterase inhibitor (placebo, n¼335; memantine, n¼342). In that trial, memantine ER treatment was associated with significant benefits compared with placebo on multiple outcome measures, including the Neuropsychiatric Inventory (NPI), a scale designed to assess behavioral symptoms in AD. Here we report results from a post hoc analysis of that trial, in which we assessed the behavioral effects memantine ER as a function of patients’ disease severity at Baseline, as determined using the Mini Mental State Examination (MMSE). Methods: Patients (observed cases; N¼540) were divided into 14 subgroups, corresponding to Baseline MMSE scores (range: 4-17). Baseline-to-Endpoint (Week 24) changes for memantine ER and placebo groups were assessed for each subgroup using a mixed-effects model with repeated measures, with the Baseline MMSE score as a linear covariate; sensitivity analyses were conducted using a quadratic model and a separate means model. Due to the exploratory nature of this analysis, no corrections for multiple hypothesis testing were performed. Results: At Week 24, memantine ER was associated with mean improvement on the NPI over placebo for patients with Baseline MMSE values across the full range of 4-17, with mean between-group differences ranging from 2.7 to 3.0 points. The mean differences reached significance (p<0.05) in the range of 7-14, with the analysis of the outlying score ranges being limited by small n values and low statistical power. Sensitivity analyses yielded similar results. Conclusions: In this post hoc analysis, 6 months of treatment with memantine ER in patients with moderate to severe AD was associated with a significant mean improvement in behavioral symptoms across a wide range of Baseline severities.