Do cognitive and neurophysiological effects of acute memantine “challenge” predict its clinical benefits in Alzheimer’s Disease?

"Personalized" interventions based on patients' "biomarkers" may be valuable for treatments that benefit only subsets of patients. The NMDA antagonist, memantine, slows clinical progression of Alzheimer's disease (AD); this effect is heterogeneous in magnitude and duration. This study tested whether acute cognitive or neurophysiological responses to memantine challenge predicted sensitivity to memantine's therapeutic effects. Thirty individuals with mild-to-moderate severity AD (M:F = 13:17) and 24 comparably aged healthy subjects (HCS) (M:F = 12:12) were enrolled. Participants with AD were characterized on 9 experimental measures and their changes after acute memantine "challenge" (20 mg). We then assessed whether acute memantine effects on these measures predicted clinical change over a 24-week open-label trial of memantine. Baseline cognitive (Repeatable Battery for the Assessment of Neuropsychological Status) and neurophysiological measures (prepulse inhibition, P3a latency and auditory steady state response power and coherence) were impaired in participants with AD (p's<0.05-0.0001); neurophysiological deficits were modestly reduced by acute memantine challenge. As a group, participants with AD showed no significant clinical changes across 24 weeks of memantine treatment; subgroups exhibited either small gains or deterioration. With one exception (mismatch negativity latency, p < 0.017), sensitivity of experimental measures to acute memantine challenge did not significantly predict clinical sensitivity to memantine. In summary, a challenge dose design identified neurophysiological measures sensitive to memantine in mild-to-moderate severity AD; acute memantine effects on one measure weakly predicted clinical outcomes over 24 weeks. Impairment in specific measures among participants with AD, and their opposition by memantine, might inform future efforts to identify treatment biomarkers in AD.