From a mouse: systematic analysis reveals limitations of experiments testing interventions in Alzheimer's disease mouse models.

The increasing prevalence of Alzheimer's disease (AD) poses a considerable socio-economic challenge. Decades of experimental research have not led to the development of effective disease modifying interventions. A deeper understanding of in vivo research might provide insights to inform future in vivo research and clinical trial design. We therefore performed a systematic review and meta-analysis of interventions tested in transgenic mouse models of AD. We searched electronically for publications testing interventions in transgenic models of AD. We extracted data for outcome, study characteristics and reported study quality and calculated summary estimates of efficacy using random effects meta-analysis. We identified 427 publications describing 357 interventions in 55 transgenic models, involving 11,118 animals in 838 experiments. Of concern, reported study quality was relatively low; fewer than one in four publications reported the blinded assessment of outcome or random allocation to group and no study reported a sample size calculation. Additionally, there were few data for any individual intervention-only 16 interventions had outcomes described in 5 or more publications. Finally, "trim and fill" analyses suggested one in seven pathological and neurobehavioural experiments remain unpublished. Given these historical weaknesses in the in vivo modelling of AD in transgenic animals and the identified risks of bias, clinical trials that are based on claims of efficacy in animals should only proceed after a detailed critical appraisal of those animal data.

Early detection of cryptic memory and glucose uptake deficits in pre-pathological APP mice.

Earlier diagnosis and treatment of Alzheimer's disease would greatly benefit from the identification of biomarkers at the prodromal stage. Using a prominent animal model of aspects of the disease, we here show using clinically relevant methodologies that very young, pre-pathological PDAPP mice, which overexpress mutant human amyloid precursor protein in the brain, exhibit two cryptic deficits that are normally undetected using standard methods of assessment. Despite learning a spatial memory task normally and displaying normal brain glucose uptake, they display faster forgetting after a long delay following performance to a criterion, together with a strong impairment of brain glucose uptake at the time of attempted memory retrieval. Preliminary observations suggest that these deficits, likely caused by an impairment in systems consolidation, could be rescued by immunotherapy with an anti-ß-amyloid antibody. Our data suggest a biomarker strategy for the early detection of ß-amyloid-related abnormalities.