Increasing fibrillar amyloidosis is associated with improved cognition and synaptic preservation caused by microglia modulation

53 Aim: Immunmodulatory treatment with the PPARγ agonist Pioglitazone shows heterogeneous results in Alzheimer`s disease (AD). We aimed to apply longitudinal in-vivo PET monitoring combined with terminal cognitive assessment to investigate PET as a tool for longitudinal monitoring of treatment effects and to establish efficacy determinants. Methods: PS2APP (TG) mice and C57Bl/6 (WT) mice were randomly stratified in Pioglitazone (N=13 TG, N=8 WT) and placebo treatment (N=10 TG, N=7 WT) at the age of 8 months. Therapy was conducted for 5 months (mo). All mice received amyloid-PET (Aβ-PET) scans (F-18-Florbetaben) at baseline and after 5 mo, to investigate progression of plaque load. TSPO-PET (F-18-GE180) was performed at baseline, after 0.5 mo, 2.5 mo and 5 mo to analyze changes in microglial activation. Forebrain PET signals were assessed for each timepoint. Behavioral testing after the terminal PET scan was followed by molecular validation including immunohistochemical (IHC) and profiling of microglia phenotypes, plaque composition and synaptic density. Results: Pioglitazone treatment resulted in a reduction of -13% (p < 0.001) in the TSPO-PET signal as well as an unexpected increase of Aβ-PET in the treatment group compared to placebo (+16% vs. +13%; p < 0.05; supporting Fig. a) for the entire monitoring period. IHC proved Aβ-PET results and showed more fibrillar plaque load in the treatment group (4.2% vs 2.8%, p < 0.05; Fig. b). Nonetheless, we observed improved cognitive performance (escape latency 3.2 vs. 4.8 seconds, p < 0.05) and attenuated synaptic loss (0.43 vs. 0.38 VGLUT1/µm2, p < 0.001) after treatment. Importantly, we discovered a direct correlation between Aβ-PET increase with TSPO decrease (R=0.62; p < 0.05) and improved cognitive outcome (R=0.39 p < 0.05). IHC revealed, that dense proportions of the Aβ plaque raised during treatment, whereas looser parts of Aβ plaques decreased. Genetic analyses indicated a treatment dependent change in microglial gene signature (Fig. c). Conclusions: Reduction of microglial activation by Pioglitazone anti-inflammatory treatment is associated with increasing fibrillar Aβ. However, this shift of loose to more dense proportions of Aβ plaques is accompanied by improved cognitive performance and attenuated synaptic loss after treatment. This finding has strong implications to Pioglitazone treatment monitoring by Aβ-PET as a signal increase over time could be misinterpreted as a treatment failure. Furthermore, this indicates for the first time that increasing fibrillar amyloidosis can preserve cognition.