Introduction: Alzheimer’s disease (AD) is pathologically characterized by deposits of amyloid-b plaques and tau protein in the brain, as well as a characteristic neurodegeneration pattern in the cortex. Current research suggests that the brain can be regarded as a functional network of different regions forming a connectome and neurodegenerative processes, as AD, can affect it. The Default Mode Network (DMN) is particularly vulnerable to AD, and its alterations can lead to the cognitive deficits characteristic of this pathology. Positron emission tomography (PET) imaging using 18F-fluorodeoxyglucose (18F-FDG), or the first minutes after administration of radiotracers targeting amyloid pathology (early-phase) can reliably assess AD characteristic patterns of neurodegeneration. Florbetapir (FBP) and Flutemetamol (FMM) are radiotracers that target cerebral amyloid deposits. The ability of early-phase amyloid imaging to assess brain connectivity has not been explored yet. The aim of this study was to assess differences in connectivity measured with 18F-FDG and early-phase amyloid imaging.
Methods: A cohort of 141 subjects from the Geneva Memory Clinic underwent 18F-FDG-PET and amyloid-PET within 12 months. Subjects were classified as Amyloid positive (A+) and Amyloid negative (A-) based on a centiloid scale threshold (12). Average uptake for Automated Anatomical Labelling atlas 3 brain regions were loaded into a graph theory software, a weighted connectivity matrix (Pearson correlation) was generated, and connectivity parameters were calculated for both 18F-FDG and early-phase images from FBP or FMM (eFBP and eFMM, respectively). Paired t-tests between networks generated by different images were performed after bootstrapping (1000 permutations).
Results: The mean age (SD) for A+ was 72.15 (5.56) and for A- 70.8 (8.29), 53.2% of the subjects were women. Changes in connectivity patterns between A- and A+ subjects were observed in the same regions for 18F-FDG and early-phase. Early-phase connectivity matrices exhibit, in general, a lower hyperconnectivity pattern compared to 18F-FDG, both for the whole brain and within the DMN. Connectivity differences were mainly observed between eFMM and 18F-FDG in A- subjects, with similar results for DMN. Assessment of modularity and average degree showed significant differences between the A- and A+ groups in eFMM (p = 0.01 for both). For differences in connectivity between eFBP and 18F-FDG, A+ subjects more frequently showed significant differences, although DMN did not always reflect these same differences.
Conclusion: Connectivity matrices derived early-phase amyloid images successfully allowed the detection of the AD changes observed in 18F-FDG PET, with less marked differences for eFMM in terms of connectivity parameters. The eFMM also showed an advantage in distinguishing connectivity profiles between the two groups of subjects, while revealing a similarity with the default mode network (DMN) for the whole brain.
