No abstract

No abstract

BackgroundSynaptic loss is a major feature of symptomatic Alzheimer’s disease (AD). PET radioligands for synaptic vesicle glycoprotein 2A (SV2A) provide an index for synaptic density in the brain. While development of these tracers is a major advance for the field of AD, little is yet known about regional synapse loss in AD or the extent to which synapse loss accounts for cognitive dysfunction in vivo. Here we tested whether regional synaptic density measured with [11C]UCB‐J is associated with cognition. Secondarily, we examined the association between tau, measured using [18F]MK‐6420 PET, and cognitive dysfunction.MethodParticipants were recruited from the Wisconsin Alzheimer’s Disease Research Center and the Wisconsin Registry for Alzheimer’s Prevention study (Table 1). Cognitive functioning was measured using the Preclinical Alzheimer Cognitive Composite (PACC)(Donohue et al, 2014). Neurofibrillary tau in the entorhinal cortex (ERC) was detected using [18F]MK‐6420 SUVR. A global measure of synaptic density was calculated using volume‐averaged DVRs in the medial temporal lobe (hippocampus, entorhinal cortex, parahippocampus, and perirhinal cortex), superior and inferior parietal cortices, and the frontal lobe. Pearson’s correlations assessed the following relationships: PACC : ERC tau, PACC : global synaptic density, and PACC : regional synaptic density.ResultERC tau and PACC score showed a significant inverse relationship: r = ‐0.51, P = 0.004. While global synaptic density and PACC score were not significantly related, when individual regions were tested, hippocampal synaptic density showed a significant positive relationship with the PACC score (P = 0.05) (Figure 1A).ConclusionERC tau PET SUVR was robustly associated with poorer cognitive performance. While synapse loss has been observed among multiple regions in AD, our preliminary results suggest that synapse loss in the hippocampus is most closely associated with cognitive dysfunction. This preliminary work suggests that ERC tau is more strongly associated with Alzheimer’s‐related cognitive decline than global or hippocampal synaptic density in participants who are mostly cognitively unimpaired. Future work will include group comparisons between participants with and without significant amyloid and tau pathology, as well as exploratory analyses into which regions are most significantly affected by Alzheimer’s‐associated synaptic loss.

Background Prior studies employing neuromelanin‐sensitive MRI have linked neurodegeneration in the locus coeruleus (LC) to Alzheimer’s disease (AD) pathology. MPnRAGE (Keckskemeti et al., 2016), a novel MRI technique developed at UW‐Madison, generates multiple inversion recovery contrasts and quantitative T1 relaxometry. Quantitative R1 (= 1/T1), the longitudinal relaxation rate, is sensitive to brain myelin, lipid, and iron content according to post‐mortem studies. As prior studies suggest altered quantitative R1 (QR1) among individuals with AD pathology, this imaging marker may be sensitive to AD‐associated pathology in the LC, one of the first regions in the brain to show signs of AD pathology (Braak et al., 2011). Here we performed preliminary assessments to test for differences in QR1 among individuals with and without clinical mild cognitive impairment (MCI‐AD), or AD dementia. We hypothesized that R1 would be lower among individuals with AD and MCI compared to unimpaired controls. Method Participants were enrolled in the Alzheimer’s Disease Connectome Project (ADCP) at UW‐Madison. 103 older adults diagnosed as cognitively unimpaired (N=53), MCI‐AD, and AD‐Dementia (Table 1) underwent MPnRAGE at a 1mm3 resolution. MPnRAGE‐derived QR1 maps were registered to MNI space using SPM12. LC masks (from Betts et al., 2017) were transformed to subject space to extract mean R1 in the bilateral LC using the Marsbar toolbox. Multiple linear regression was run within Rstudio to investigate the relationships between age, R1, and diagnosis. The main effects of interest included age and diagnosis, as well as their interaction. Sex was included as a covariate. Result No significant associations were found between QR1 and age, diagnostic status, and their interaction. The main effect of age on R1 trended toward significance (p=0.059). Age trends are shown in Figure 1. Sex was not a significant covariate. Conclusion While QR1 was not significantly associated with age or clinical diagnosis within the LC, further research will evaluate the extent to which QR1 is altered among other regions affected by AD pathology. Further, R1 could be tested between groups stratified by AD biomarker status. Finally, the LC is a small structure, thus future work could also test R1 estimated at higher resolution.

No abstract