Longitudinal multimodal imaging in mild to moderate Alzheimer disease: a pilot study with memantine

Abstract

Objective: To study the feasibility of multimodal neuroimaging in mild to moderate Alzheimer disease (AD) and to estimate the size of possible treatment effects of memantine on potential functional, structural and metabolic biomarkers of disease progression.

Methods: In this randomised, double-blind, placebo-controlled pilot study, 36 patients with moderate AD received 52 weeks of memantine (20 mg/day) or placebo. Patients were re-evaluated after 26 and 52 weeks to measure the change from baseline in several outcome measures including global and regional glucose metabolism, total brain and hippocampal volumes, as well as chemical shift imaging-derived global and regional N-acetylaspartate and myoinositol concentrations.

Results: In the total population, global glucose metabolism decreased by 2.3% (p<0.01), total brain volume by 2.1% (p<0.001) and hippocampal volume by 2.7% (p<0.01) after 52 weeks. Chemical shift imaging (CSI) spectra were severely affected by patient-induced artefacts and highly variable. Patients receiving memantine showed less decline in glucose metabolism in all brain areas than patients on placebo. Their loss of hippocampal volume was substantially smaller (2.4% vs 4.0%). No between-group differences were seen for changes in total brain volume.

Conclusions: The results support the use of multimodal imaging including MRI and positron emission tomography (PET) to monitor the progression of moderate AD. CSI yielded unreliable longitudinal results. The data suggest that memantine has potentially protective effects in AD and they can be used for planning larger confirmatory studies on the cerebral effects of memantine.

Figure 1. A–D. Baseline (A) and 52-week (B) positron emission tomography (PET) scans registered on MRI in an 85-year-old female study participant who experienced a focal, almost complete loss of glucose utilisation in the right temporoparietal region. Glucose metabolism in other brain regions remained almost unchanged from baseline. The baseline scan demonstrated the typical symmetric temporo-parietal hypometabolism of patients with Alzheimer disease (AD). The coregistered T₁-weighted MRI scans show little, if any, sulcal enlargement in the area of metabolic loss in this patient between baseline (C) and follow-up (D).

Figure 2. Regional and global changes from baseline in glucose metabolism in patients with Alzheimer disease in a 52-week randomised, double-blind, placebo-controlled trial. Numbers below regions indicate the percentage decline in patients receiving memantine vs those receiving placebo. After 52 weeks, memantine-treated subjects had smaller reductions in all brain regions and globally. Differences did not reach statistical significance in this pilot study.