Relationship between Neuroimaging and Cognition in Frontotemporal Dementia: A [18 F]FDG PET and Structural MRI Study.

Background: Frontotemporal dementia (FTD) is a clinically and pathologically heterogeneous condition with a prevalence comparable to Alzheimer's Disease for patients under sixty-five years of age. Gray matter (GM) atrophy and glucose hypometabolism are important biomarkers for the diagnosis and evaluation of disease progression in FTD. However, limited studies have systematically examined the association between cognition and neuroimaging in FTD using different imaging modalities in the same patient group.

Methods: We examined the association of cognition using Montreal Cognitive Assessment (MoCA) with both GM volume and glucose metabolism using structural magnetic resonance imaging (MRI) and F-fluorodeoxyglucose positron emission tomography scanning ([F]FDG PET) in 21 patients diagnosed with FTD. Standardized uptake value ratio () using the brainstem as a reference region was the primary outcome measure for [F]FDG PET. Partial volume correction was applied to PET data to account for disease-related atrophy.

Results: Significant positive associations were found between whole-cortex GM volume and MoCA scores (r = 0.461, p = 0.035). The association between whole-cortex [F]FDG and MoCA scores was not Significant (r = 0.374, p = 0.094). GM volumes of the frontal cortex (r = 0.540, p = 0.011), caudate (r = 0.616, p = 0.002), and insula (r = 0.568, p = 0.007) were also Significantly correlated with MoCA, as were values of the insula (r = 0.508, p = 0.018), thalamus (r = 0.478, p = 0.028), and posterior cingulate cortex (PCC) (r = 0.472, p = 0.030).

Discussion: Whole-cortex atrophy is associated with cognitive dysfunction, and this effect is larger than for cortical hypometabolism as measured with [F]FDG PET. At the regional level, focal atrophy and/or hypometabolism in the frontal lobe, insula, PCC, thalamus, and caudate seem to imply the importance of these regions for the decline of cognitive function in FTD. Furthermore, these results highlight how functional and structural changes may not overlap and might contribute to cognitive dysfunction in FTD in different ways. Our findings provide insight into the relationships between structural, metabolic, and cognitive changes due to FTD.